January 31 News
Healthcare: Nanning First People’s Hospital (南宁市第一人民医院) deployed an intelligent guidance service to help patients register more quickly, then expanded into a “digital human doctor” launched in February 2025 to provide automated consultation-style Q&A alongside voice-recognition AI inquiry and an AI medical guidance system, positioning the virtual clinician as a front-door triage and navigation layer rather than a replacement for in-person care; Guangxi Medical University First Affiliated Hospital (广西医科大学第一附属医院) announced an interactive medical specialty digital human agent called “Mi Yu Yi Tan” (泌语医谈), presented as a case study of hospital-grade digital-human delivery that specifically targets real-time interaction and supports ASEAN minority languages, framing the virtual specialist as a multilingual interface for specialty information exchange and guided consultation flows in cross-border or regional care contexts.
Education: Guangxi Party Member Education Center (自治区党员教育中心) and the Nanning Municipal Party Committee Organization Department (南宁市委组织部) staged an immersive “party class” experience in which technical staff used digital-human methods to render “Huang Lao” (黄老) as a digital persona that could conduct remote, dialogue-style interaction with trainees, using the virtual figure as the core instructional device to create a “time-separated” conversational learning scene; China Unicom (中国联通) applied its digital-human generation capability to produce promotional posters for finalists in an Anhui university singing competition, using synthetic human imagery as the standardized visual production pipeline for campaign assets rather than commissioning individual photography or illustration.
Culture And Heritage: Wudang Mountain (武当山) in Hubei Province launched the "入境武当" (Entering Wudang) VR large-space experience as part of its smart tourism upgrade, featuring virtual tour guides named "玄淼淼" (Xuan Miaomiao) and "武焱焱" (Wu Yanyan), developed by Wuhan Dashew Technology Co., Ltd. (武汉大势智慧科技有限公司) in partnership with Wudang Mountain Cultural Tourism Development Group (武当山文旅集团), utilizing PICO VR hardware within the Wudang Dream park complex. Shandong Province's cultural and tourism sector has deployed multiple AI-enhanced tools including scenic-area tour guides and Confucius-themed digital humans; the Confucius digital human (孔子数字人) and its underlying Confucius intelligence large model (孔子数智大模型) were jointly developed by Shandong Digital Culture Group (山东数字文化集团) in partnership with People's Daily Online (人民网), Shandong Academy of Social Sciences (山东社会科学院), Nishan World Confucianism Center (尼山世界儒学中心), Shandong University Confucian Studies Institute (山东大学儒学高等研究院), and related provincial entities, while the Qufu Three Confucius Sites (曲阜三孔) scenic area's AI tour guide system was created by Shanyan Digital Technology (Shandong) Co., Ltd. (山演数字科技(山东)有限责任公司), which specializes in AR and AI navigation solutions for cultural heritage destinations and officially launched the system in March 2024.
Commerce And Marketing: JD.com (京东) positioned “AI-native digital humans” as a holiday shopping livestream workforce during its New Year goods festival, describing them as fully AI-generated in appearance, voice, and motion under the JD JoyAI large model (JoyAI大模型), and framing the capability as a scalable method for producing livestream presenters across different persona styles; Tiandi Online (天地在线) was discussed in the context of an “AI digital human + XR livestream + digital content production” service chain, describing a commercially packaged workflow where digital humans are used as the on-camera talent layer while XR scenes and content production complete the delivery loop for marketing and sales broadcasts; Hangzhou cross-border e-commerce reporting described a promotional “AI digital human + virtual scene” livestream solution aimed at multilingual selling, treating the virtual presenter and synthetic environment as an integrated export-facing merchandising format rather than a novelty demo.
Governance And Legal: Beijing governance coverage cited disputes and oversight concerns centered on AI voice-based selling and virtual digital-human rights, including a Beijing Internet Court determination that a disputed virtual human image qualified as an artwork and that a platform-side company (named only as “Xi Mou Company,” 西某公司) bore joint liability in an infringement case, illustrating how platform distribution and model-based image generation can translate into concrete civil exposure; Jiangsu policy reporting described plans to set up an enterprise liaison-point system for “digital-human livestream OPC” firms and to publish compliance guidelines explicitly aimed at controlling marketing and consumer-risk issues associated with digital-human livestreaming; Inner Mongolia legal-services reporting presented “digital human lawyers” as a service modality intended to integrate with human lawyers, positioning the synthetic legal persona as a front-end interaction and delivery layer for public and enterprise legal assistance within a “technology + governance” framework.
Media And Communications: Hong Kong Wen Wei Po (香港文匯報) presented the use of an AI digital news anchor named “Chen Dawen” (陳大文) to deliver detailed event coverage as part of an awards-style announcement format, treating the virtual anchor as a production component for scripted, structured reporting rather than a standalone personality-driven channel; a criminal-case report noted that its accompanying video content used the Huiboxing digital human (慧播星数字人), indicating a newsroom-style pipeline in which a synthetic presenter is used for narration and presentation in place of a human on-camera host.
Research And Core Technology: ByteDance (字节跳动) was reported to have released FlowAct-R1 as a real-time interactive virtual digital-human technique framed around responsive interaction rather than pre-rendered performance, while explicitly acknowledging misuse risk for fake or harmful content and presenting access-control commitments as part of responsible deployment; Kunlun Tech (昆仑万维) open-sourced SkyReels-V3 as a multi-modal video-generation stack that bundles three concrete capabilities—reference-image-to-video generation, video extension, and audio-driven talking-avatar output—placing “virtual character” generation (including digital-human-style talking avatars) inside a single architecture meant to support practical content workflows from still input through speech-driven performance; Baidu (百度) was described as using digital humans as a public-facing channel to respond to commercialization and deployment skepticism, contextualized within its broader “chip–cloud–model–application” full-stack narrative, positioning the digital human as a communication interface for productization claims rather than merely a demonstration asset.
January 30 News
Tourism And Cultural Heritage: Chongqing policy discussion framed digital humans as part of an AI-enabled tourism toolchain alongside “AI companion touring” and intelligent translation devices, positioning virtual guides and on-trip assistance as a concrete application for visitors in cultural and travel settings. Bangyan Technology (邦彦技术股份有限公司) presented its NuwaAI digital human “Gankeng Li Bai” in a Harbin ice-and-snow tourism promotion setting tied to Shenzhen cultural tourism, including a cross-regional live interaction in which the digital character exchanged content with a Harbin local culture communicator around festival customs, city life, and cultural expression, treating the digital human as a real-time cultural interlocutor rather than a static mascot. Tiandi Online (天地在线) was described as expanding VR cultural-tourism projects in collaboration with its Hainan state-owned shareholder Lingfa Investment (陵发投资), while also extending AI digital humans into adjacent immersive formats including XR live streaming and LBE “space cinema,” linking digital humans to experiential tourism delivery rather than only online content.
Government And Public Services: Shaanxi Radio & Television Network (广电网络) was described as using “Jiufang Digital Human” capabilities and AIGC-oriented components as part of a broader digital-government push, and specifically deploying a government-affairs AI agent called “Monthly Report Assistant” into Xi’an, emphasizing a replicable government services model built on AI rather than a one-off demo. Hangzhou’s “12345” service channel was described as creating a hotline AI system “Lingguang Xiaohang” that operates in a digital-human form to provide real-time-updated “government value-added services,” built through collaboration between the hotline platform and Hangzhou Daily Press Group’s “Orange Persimmon Interactive” (杭报集团“橙柿互动), treating a digital human as the front-end interface for civic inquiry and service guidance. In Qinghai, Xining’s “digital-intelligence business district” was described as using digital-human shopping guides and AI shopping guides as a visible consumer-facing layer of the local digital economy narrative, placing avatars/digital humans directly into retail wayfinding and product guidance rather than limiting them to back-office automation.
Media, Marketing, And Commerce: Straits Metropolis Daily Press (海峡都市报社) described a “Million Digital Humans Plan” that moved from internal newsroom scenarios to external government-and-business services and then toward “digital human + large model” industry applications, framing digital humans as reusable production-and-service assets that can graduate from editorial operations to paid B2B scenarios. Heze Vocational College (菏泽职业学院) was described as using “digital human livestream” techniques to build a combined “online + offline, digital human + livestream” sales matrix for products tied to intangible cultural heritage contexts, treating a synthetic presenter as part of an e-commerce conversion pipeline. Jiangsu policy discussion referenced establishing an “OPC enterprise contact-point system” for digital-human livestreaming companies and issuing related rules, treating digital-human commerce as an administratively governed business category rather than an informal creator practice.
Technology Systems And Platforms: ByteDance (字节跳动) was reported as releasing FlowAct-R1, a real-time interactive virtual digital-human generation system designed for natural conversation with users, with a technical framing that emphasized real-time audiovisual performance characteristics such as lip-sync, facial expression, and gesture alongside a memory-bank mechanism and a diffusion-based strategy intended to maintain interactive fidelity. Sun Yat-sen University (中山大学) was referenced in connection with an authorization listing for a face-video generation patent oriented to improving lip-sync accuracy between virtual-human mouth movements and speech, and the ACTalker system was described as having early landing in gaming and social scenarios within a Tencent ecosystem context, positioning multi-modal inputs as a route to generating speaking-character video consistent with a target role’s appearance rather than relying solely on scripted animation.
Cultural And Ethical Contexts: A Beijing commentary on the film or drama “Extreme Judgment” described a fictional “digital human judge” named Maddox, explicitly using an AI-investigation premise as narrative scaffolding and framing the digital-human judicial role as an AI-era cautionary device rather than a deployment report, which nevertheless serves as a concrete case of how digital-human authority figures are being imagined in mainstream storytelling. A Henan “digital proposal” case described an AI-generated, visually distinct form of political-cultural communication in which APUS (APUSAI) generated a proposal artifact that ran from cultural-material collection to content generation and then to “digital human performance,” tying digital humans to the staging and presentation of heritage narratives such as digitized-and-modeled oracle-bone inscriptions, virtual-space restoration of Longmen Grottoes statues, and digital-human-based renditions of Shaolin kung fu, Henan opera, and Yellow River culture for international marketing. A Gansu proposal case described using AI and metaverse techniques for a Lanzhou beef-noodle industry scenario by piloting a “beef-noodle metaverse experience hall” in core commercial districts that combines digital-human interaction with visualization of the production process and links online activities into a closed-loop consumption pattern, along with brand virtual IP creation and digital-collectible development as the surrounding commercialization layer.
January 29 News
Cultural & Tourism: A city-level official virtual digital human “Xiao Jinhua” was built for Dali in Yunnan as an example of a government-backed tourism-facing character intended to represent a locality and support public-facing explanation and guidance; in Shanxi, Shiyou Technology (世优科技) created the Chinese-style digital human “Qingniao” by extracting visual inspiration from the “bird” motif on the Shanxi Museum’s signature artifact “Jinhou Bird Zun,” then deploying the character in major cultural-and-tourism events such as the provincial tourism development conference to serve as a youth-oriented cultural symbol; in another heritage-and-experience setting, a digital-human IP “Xin Xiaozhui” was integrated into an industrial-tourism and traditional Chinese medicine culture experience that also used immersive carriers such as a 5D flying theater, positioning the synthetic character as a recurring narrative guide within a broader experiential venue design.
Government & Public Services: Nanjing’s 12345 hotline expanded multi-channel access by rolling out the “Ning Suban” online service zone (宁诉办网络服务专区), jointly advanced by Nanjing Big Data Group (南京大数据集团有限公司) and the Nanjing “12345” Government Hotline Service Center (南京市“12345”政务热线服务中心), adding a DeepSeek-model digital-human customer-service function described as 24/7, high-concurrency, second-level response, and human–AI collaborative routing across government-service topics via channels including “My Nanjing” (我的南京), “Ningqi Tong” (宁企通), and official WeChat/mini-program entry points. Tianjin Bank (天津银行股份有限公司) disclosed procurement results for an online customer-service upgrade that adds a digital-human layer plus an “intelligent Siri” capability, with Beijing Zhongke Jincai Technology Co., Ltd. (北京中科金财科技股份有限公司) named as the winning bidder in the published award notice. In Hainan’s Tunchang prosecutorial outreach, the “Tun Daxia” (屯大侠) stylized character set (with anti-bullying, anti-drug, and anti-fraud companion personas) was paired with an AI digital human “Jian Xiaofan” (检小凡) generated from a staff member’s likeness for internet-native legal-education messaging, but public reporting does not name an external vendor. In Chengdu, the interactive digital human “Rong Xiaobu” (蓉小布) was presented as an on-site public-facing interface jointly built by Chengdu Release (成都发布) and Chengdu Mobile (成都移动), described as meeting-linked Q&A drawing on authoritative information sources.
Enterprise & Industry: Supcon Technology (中控技术) used a virtual host “Zhi Xiaokong” in a large annual corporate event context, treating a synthetic on-stage presenter as a branded interface for industrial-AI messaging and internal stakeholder engagement; Jishi Media (吉视传媒) reported deploying digital-human solutions in major cultural events (including a heritage conference setting) while also referencing a “Yumi” AI large model as part of its broader media-technology portfolio, linking synthetic presenters and event-facing digital characters to platform-level AI capabilities intended for media production and distribution workflows.
Marketing & E-commerce: JD.com (京东) was described as scaling “AI-native” digital humans for livestream commerce, framing digital-human anchors as a production and conversion engine that can operate continuously and normalize synthetic hosts within major retail promotional cycles; Silicon Intelligence (硅基智能) was cited as providing digital-human livestream hosts capable of 7×24 operation, emphasizing persistence and throughput as the commercial value proposition for synthetic presenters used in sales-driven broadcasting; Bangyan Technology (邦彦技术股份有限公司) deployed a “NuwaAI” digital human as a representative figure in a cross-regional “technology + culture-and-tourism” promotion setting connected to Harbin’s ice-and-snow festival programming, using the synthetic character as a marketing-facing emissary to support destination branding and themed tourism promotion.
Entertainment & Social Platforms: Tencent (腾讯) renewed QQ Show as an AI-driven virtual-avatar system that can substitute for static profile images by entering chat scenes directly, positioning AI-generated virtual appearances as a core element of social expression and interaction within messaging; Baidu (百度) was described as embedding an always-available “Wenxin Assistant” concept into group-chat participation and positioning it as a personal “digital alter” that supports practical tasks such as travel planning and weather lookup inside the chat context, treating the assistant-as-avatar pattern as a default companion layer for each participant rather than a separate, opt-in tool.
Legal & Regulatory: The Ministry of Industry and Information Technology of the People’s Republic of China (中华人民共和国工业和信息化部) initiated public consultation on a mandatory national-standard project for “digital human identity identifiers” under a metaverse classification-and-identification framework, specifying a system for identifier structure plus registration, management, and marking requirements that implies a “one human, one code” approach intended to standardize identity signaling for digital humans across deployment contexts; in Beijing, court and procuratorate reporting explicitly surfaced disputes involving AI-generated voice selling and infringement of virtual digital human images, treating synthetic voice commercialization and digital-human likeness rights as concrete enforcement domains alongside other AI-era competition and data-governance case types.
January 28 News
Government And Civic Services: At the Chengdu “Two Sessions” venue in Sichuan, a dedicated interview and interaction zone run by Chengdu Media Group (成都传媒集团) became a popular onsite “check-in” area where attendees could directly experience a digital human and related interactive cultural experiences, including a hands-on trial of the Chengdu city digital human “Rong Xiaobu” (蓉小布) by a municipal political consultative conference delegate from Chengdu Guancheng Group (成都冠城集团) who also serves as an executive director of Chengdu Guancheng Experimental School (成都冠城实验学校), positioning the digital human as an on-site civic engagement interface rather than a purely online service. In Nanjing, the city’s 12345 public hotline expanded its multi-channel service system by launching a DeepSeek-based digital human intelligent customer-service capability inside the “Ningsuban” (宁诉办) online service area, emphasizing 24/7 availability, high concurrency handling, second-level response speed, and human–AI collaborative routing for issues handled through the “My Nanjing” (我的南京) service ecosystem, making the digital human a front-end triage and interaction layer for municipal requests.
Marketing: In large-scale commerce livestreaming, JD.com (京东) was reported as pushing “AI-native” digital humans into broader commercialization, including holiday shopping scenarios, with its approach described as being powered by the JoyAI large model (JoyAI大模型) and framed around moving livestreaming into an “AI-native” operational mode; the same reporting described themed digital-human content including historically styled “ancient” digital characters used for promotional engagement, indicating a shift from limited pilots to repeated, campaign-based deployment. In regional brand marketing, Tianyoude Liquor (天佑德酒) described integrating a branded digital human with Qinghai’s regional characteristics and the historical-cultural narrative of Qingke barley liquor, explicitly treating the digital human as a recognizable identity carrier that merges traditional cultural elements with digital presentation for differentiated product storytelling.
Education And Training: In Wuhan, the Wuhan Federation of Trade Unions (武汉市总工会) partnered with Baidu (百度) to offer no-cost public training that explicitly included “digital human applications” alongside content creation and ecommerce livestreaming, treating digital humans as a practical employability skill within a broader applied-AI curriculum rather than as a specialist R&D topic. In Zhengzhou, an ideological-and-political education observation activity in the Zhengdong New District included classroom use of an AI digital human as part of a lesson themed on the spirit of China’s crewed spaceflight program, showing the digital human being used as an instructional aid embedded in a standard teaching demonstration context rather than as a standalone product showcase. In tertiary and professional education settings in Guangxi, a 3D digital human interface was reported as the interaction surface through which students queried ASEAN legal knowledge, positioning a digital human avatar as the front-end for structured legal-information access in a classroom environment.
Legal And Regulatory: Beijing judicial reporting explicitly incorporated multiple AI-and-digital-human disputes into official work-report coverage, including cases involving AI-generated voice used for livestream selling, infringement claims involving a virtual digital human image, disputes characterized as unfair competition linked to “comic transformation” effects, and matters involving illegal scraping of aggregated data from online platforms, collectively indicating that synthetic media and digital-human representations are now being treated as recurring fact patterns across consumer commerce, IP-style image rights, and platform data governance. At the national policy level, the Ministry of Industry and Information Technology (工业和信息化部) was reported as proposing the development of standards to规范 digital-human identity identifiers, with the shorthand goal described as enabling a “one person, one code” (一人一码) mechanism for digital-human identity marking, signalling a move toward standardized identification and traceability for deployed digital-human entities.
Enterprise: In Beijing’s “data elements × AI” practice, China Telecom Beijing (北京电信) and the Mentougou District Government Service and Data Administration (门头沟区政数局), together with Beijing University of Posts and Telecommunications (北京邮电大学), were reported as jointly building an industry-oriented data intelligent labeling innovation center that includes “digital human ecommerce” among its focus areas, indicating an enterprise-and-government data pipeline designed to support production-grade digital-human commercial operations via structured annotation and operational tooling. In platform development, the Xingchen Intelligent Agent Platform (星辰智能体平台) was reported as upgrading toward multimodal interaction and explicitly framing a digital human as a warmer “emotional companion” interaction role, positioning the digital human as the user-facing embodiment of an agent platform rather than a separate, static avatar asset.
Culture: In museums and cultural display spaces, Juyou Technology (巨有科技) was reported as delivering a “digital museum” experience centered on an AI digital human guide, with concrete functions described as multilingual explanation and cultural-object narrative performance, placing the digital human in a curator/guide role intended to deepen immersion and educational engagement. In Zhejiang’s safety-service context, Hangzhou City Brain Co., Ltd. (杭州城市大脑有限公司) was reported as launching a security-service digital human named “An Xiaolong” (安小龙) to support ordinary engineers, indicating a specialized digital human persona designed to act as an operational assistant within cybersecurity or safety-service workflows rather than as a general-purpose public-facing mascot. In media and broadcasting practice in Shandong, a provincial outlet’s “digital human broadcast” format was described as part of a recurring reporting series, reflecting the use of a digital human presenter as a packaging layer for news delivery rather than a one-off stunt.
Technology And IP: On the technology supply side, Wuxi Shunyi Intelligent Technology Co., Ltd. (无锡顺熠智能科技有限公司) was reported as filing a patent for a digital-human image generation system based on personalized feature modeling, explicitly linking its technical approach to improved personalization capability and user experience in generated digital-human representations. On the business side of the sector, Shuziren Technology (山东数字人科技股份有限公司) issued a 2025 performance forecast indicating an expected return to profitability with net profit attributable to shareholders projected at RMB 7.5–8.5 million, reflecting ongoing commercialization and financial reporting activity in the domestic digital-human industry even when the reporting focus is corporate results rather than a single deployment case.
January 27 News
Governance: China’s Ministry of Industry and Information Technology (工业和信息化部) initiated public consultation on a compulsory national standard project titled “Metaverse Classification and Identification: Requirements for Digital Human Identity Identification” (《元宇宙分类与标识数字人身份标识要求》), positioning “digital human identity identifiers” as a standardized, regulated layer of metaverse classification and labeling, with reporting framed as moving toward an “one person, one code” style of identity marking for digital humans and related governance of identity attribution across digital-human deployments.
Marketing: Baidu (百度) is described as advancing digital-human livestream commerce as a concrete growth case, including performance claims for a flagship digital-human debut livestream exceeding 13 million cumulative views and approximately RMB 55 million GMV, with some core categories reportedly surpassing human livestreamers in sell-through; within the same reporting cluster, Baidu’s operational move to consolidate its digital-human business is characterized by redeploying roughly 15 digital-human team members into its e-commerce business line, and provincial risk-control responses are described in parallel policy language, including Jiangsu’s plan to establish an “OPC enterprise contact-point” mechanism for digital-human livestreaming and to issue a digital-human livestream compliance guideline aimed at preventing marketing and compliance risks in digital-human-driven selling and promotion.
Cultural: Juyou Technology (巨有科技) is presented through a museum case study in Jiangxi in which an AI digital human named “Huanhuan” (环环), modeled on a Shang-dynasty “Feathered Person Jade Pendant” (商代羽人玉佩饰), delivers child-friendly artifact interpretation, using dynamic motion-capture-driven movement plus affective/emotion-computing techniques, and enabling visitors to trigger 3D animated content via AR scanning to visualize an “animated” rendition of the character and related craftsmanship narratives as part of an immersive cultural explanation workflow.
Entertainment: Zhejiang Television (浙江卫视) and China Mobile Zhejiang Research Institute (中国移动（浙江）研究院) are cited as jointly developing an “AI + XR large-space” experience via Zhejiang TV’s digital-human studio collaboration model, situating digital humans as real-time or semi-real-time presentational entities inside extended-reality venues and event programming; separate exhibition-oriented reporting also describes Hangzhou Arts and Crafts Museum integrating multiple original animation IP characters (“Yujian,” “Huasen,” “Nishan,” and “Jiangbao,” 御剑、婳伞、霓扇、匠宝) into on-site presentation in the form of digital humans and 3D animation, treating synthetic characters as exhibition narrators and brandable interpretive figures rather than purely decorative visuals.
Enterprise: Zhewen Internet (浙文互联) is referenced in a listed-company disclosure context that explicitly bounds its “Paizhi” (派智) digital-human business as contributing only a small share of revenue and profit (reported as below 1%), framing the digital-human line as commercially present but not yet material to overall financial performance; separately, Silicon Intelligence (硅基智能) is associated with Nanjing’s “AI Time-Space” community narrative through a “digital avatar/clone” style capability described as “The Art of Doppelganger” (分身有术), positioning digital-human “split-body” style embodiments as part of an applied city-level AI showcase rather than a purely experimental lab artifact.
January 26 News
Governance: Ministry of Industry and Information Technology (工业和信息化部) initiated a mandatory national standard project for “Metaverse Classification and Identification—Digital Human Identity Identification Requirements” (《元宇宙分类与标识数字人身份标识要求》), framed around closing gaps in digital-human identity management by specifying identity-identification requirements that operationalize an “one person, one code” approach, with the policy motivation explicitly linked to misuse scenarios such as identity theft or imitation of someone else’s digital-human likeness to create and disseminate illegal or harmful content, reputational harm via impersonated digital humans, and technical governance concerns including insufficient content filtering and weak control points around identity attribution for synthetic-persona outputs.
Entertainment: Tencent (腾讯) deployed character-based AI “clones” that remain persistently available to audiences for interactive dialogue while watching drama content, built on the Hunyuan model (混元大模型) and implemented as always-on role instances for named protagonists “Lin Yusen” (林屿森) and “Nie Xiguang” (聂曦光), with the interaction design described as script-ingested role comprehension, reconstruction of key scenes, and in-character conversation from the role’s viewpoint; in parallel, it revived QQ Show (QQ秀) as an AI-driven avatar system that can generate a virtual image and embed it directly into chat so the avatar can substitute for a traditional profile picture inside conversational contexts rather than remain a static identifier.
Healthcare: Ant Group (蚂蚁集团) scaled “AI doctor clones” inside Afu App (阿福App) by enabling large numbers of clinicians to run personal AI instances that handle high-volume, routine questions before escalation to human clinical judgment, with reported deployment metrics including more than 1,000 participating doctors and roughly 27 million accumulated consultations handled through these AI instances; it also described an “AI assistant” workflow that supports clinical operations by pre-visit record capture and pre-triage style questioning to reduce repetitive intake burden, and the broader service integration was contextualized as accelerating collaboration after its acquisition of Haodf Online (好大夫在线).
Patents And R&D: Beijing Xiaochuan Online Network Technology (北京小川在线网络技术有限公司) filed a patent application for a virtual-human online chat method that generates anthropomorphic replies by leveraging “memory conflict” as a mechanism, described as an approach to produce more human-like conversational outputs when a virtual being’s stored memories contain contradictions or competing cues, with the filing published as CN121388100A; separately, a long-term-memory benchmark for AI clones was discussed as a way to evaluate whether a digital “clone” can preserve details and temporal structure rather than compressing user history, explicitly tied to near-term deployment targets for AI digital doubles such as email drafting, schedule management, and acting as a persistent interlocutor across generations.
Creator Tools And Production: Wondershare Technology (万兴科技) was presented as combining an audio-video foundation model branded as “Wanxing Tianmu” (万兴天幕) with 3D figurine-style effects and AI digital-human capabilities for short-video creation, 3D effect production, and virtual-human live streaming, including a specifically named “3D figurine effect” tool positioned as an end-user feature for stylized outputs; Cloudwalk Technology (云从科技) was described as building virtual digital-human capabilities in tandem with “digital twin” and “metaverse” positioning, with commercialization targets spanning virtual live broadcasting, virtual customer service, and animation/game scenarios; Bangyan Technology (邦彦技术) was described as advancing a product called NuwaAI digital human that combines an “EQ brain + IQ brain” architecture and is positioned to collaborate with a cloud PC environment, with the emphasis placed on pairing affective-style interaction capacity with task/problem-solving capability in a single synthetic persona.
Enterprise And Ecommerce Operations: Baidu (百度) was described as reorganizing its digital-human team into its ecommerce business line under the leadership of vice president Ping Xiaoli (平晓黎), including the movement of roughly 15 staff from the search organization into the ecommerce-aligned digital-human team, indicating a shift toward operationalizing digital humans as ecommerce-facing assets rather than keeping them primarily as search-adjacent R&D, and framing “digital human” development as a cross-business capability that can be reassigned to the business unit expected to monetize it most directly.
January 25 News
Enterprise And Government: Chongqing Municipal Human Resources and Social Security Bureau (重庆市人力资源和社会保障局) was reported as having built the “145N” digital human-resources-and-social-security system, with core business digitization reaching 100%, framing “AI + high-quality employment” as an integrated operational scenario for public services that supports large-scale employment targets in Chongqing, and positioning digital-human-style service interfaces and end-to-end digitized workflows as the concrete implementation mechanism inside a municipal labor and social security stack rather than a standalone media product; in a separate enterprise-facing case, Baidu (百度) was reported to be advancing a group-wide consolidation of its digital-human work by moving some search personnel into livestream and video teams, and its marketing-event ecosystem in Zhengzhou was described as using an AIGC creative platform called “擎舵” together with digital humans as packaged capabilities offered to local businesses for brand and operations growth, indicating an internal reallocation toward video-centric production and commerce-facing deployment of synthetic presenters and sales avatars.
Marketing And Commerce: JD.com (京东) was described as using digital humans to upgrade home-appliance and home-furnishing marketing through tightly coupled “real-person + digital-human” co-performance, AIGC-assisted content production, and digital-human hosting formats that expand event-style retail programming, while a Hong Kong-facing report quantified platform outcomes from the previous “Double 11” period as digital-human livestream gross transaction value exceeding RMB 2.4 billion and about 50,000 merchants participating via digital-human livestreaming, framing digital humans as an adoption lever for small and mid-sized sellers who want standardized livestream labor, repeatable scripts, and scalable hosting rather than bespoke influencer deals; separately, BlueFocus Intelligent Communications Group (蓝色光标) was referenced via its “蓝标智播” virtual digital-human matrix as an applied package for brand virtual livestreaming and IP derivation, including a cited case pattern of building virtual spokespersons for automakers and running high-frequency livestream sessions at scale, treating synthetic hosts as a managed service with operational throughput as the performance metric.
Culture And Entertainment: ByteDance Volcano Engine (字节跳动火山引擎) was reported as embedding AI into Spring Festival Gala production across program making, online interaction, and livestreaming, including virtual hosts appearing alongside human hosts and AI-generated stage visuals, presenting the digital-human host as a culturally mainstream broadcast device tied to real-time audience interaction and large-scale live production workflows; Tencent (腾讯) was reported to be reviving QQ Show as an avatar system that can replace static profile photos inside chat contexts, with the update described as applying AI in the character-creation step to lower the barrier to making individualized virtual images, positioning a legacy Chinese avatar product as a contemporary “avatar-as-interface” layer inside messaging rather than a nostalgic cosmetic feature; alongside these platform-led cases, a separate commentary on “virtual person” operations highlighted a concrete failure mode in which performers (“中之人”) without systematic training or professional support struggle to keep the performer identity decoupled from the character (“皮套和真人解绑”), leading to public incidents where personal mistakes or off-script speech spill over and damage the virtual character brand, treating operational governance and professionalization as the practical constraint on virtual-character sustainability.
Tourism, Transport, And Urban Services: Haikou Maritime Safety Administration (海口海事局) was reported as deploying an intelligent voice digital human called “小海狮” to improve efficiency in Qiongzhou Strait passenger-and-freight operations during the first month of Hainan Free Trade Port “closed operation,” presenting a concrete transport-administration use case where a named synthetic voice agent is used as a service interface to increase throughput per interaction; Ministry of Culture and Tourism of the People’s Republic of China (中华人民共和国文化和旅游部) reporting on the Anhui Provincial Department of Culture and Tourism’s creative-sharing event described multi-organization demonstrations around “AI empowering cultural-and-tourism content creation,” explicitly including AIGC, multimodal large models, and digital humans as production tools for tourism-facing creative output, indicating an institutional push to normalize digital-human presenters and synthetic performers inside regional tourism content pipelines rather than treating them as experimental marketing stunts.
Industry Infrastructure And On-Site Interaction: MoFa Technology (魔珐科技) was described as commercializing large-screen AI digital humans across multiple scenarios, contrasting earlier “animated sticker” virtual images with current AI digital-human hosts that combine “intelligence” (AI-driven understanding and generation) with more capable visual presentation to make public-facing screens act like interactive service entrances, and positioning the product category as a next-generation human–machine interaction gateway for offline spaces; Zhongyuan Environmental Protection Co., Ltd. (中原环保股份有限公司) was reported to have obtained a patent for a “digital human interaction management system” based on big-data analysis, indicating an enterprise software pattern where digital-human interactions are treated as managed, analyzable sessions with governance and optimization logic rather than simple front-end animation; Hangzhou Shuluan Cloud Technology Co., Ltd. (杭州数峦云科技有限公司) was described in the context of a “digital twin pipeline network platform” solution that deeply integrates IoT, BIM, GIS, digital humans, XR, 3D, and AI, presenting a concrete infrastructure-stack example where digital humans are explicitly one layer inside an operational digital-twin platform for utilities-style network management rather than an isolated customer-service bot.
Standards And Regulatory Contexts: Ministry of Industry and Information Technology of the People’s Republic of China (中华人民共和国工业和信息化部) was reported to be soliciting public意见 on a mandatory national standard project titled “Metaverse Classification and Identification—Digital Human Identity Identification Requirements,” with the core practical target described as implementing “one person, one code” identity labeling for digital humans to reduce security risks, framing digital-human deployment as an identity-and-traceability problem that requires standardized classification and identifiers in metaverse-style scenarios; in parallel, platform governance was referenced through Douyin (抖音) in relation to China’s AI-generated synthetic content labeling requirements, described as involving both visible labels and hidden metadata and extending to platform review obligations for real-name registration of virtual humans, tying digital-human production and operation to compliance workflows that treat provenance marking and account identity as enforceable controls rather than optional best practices; separately, local industrial-policy discourse in Wuhan’s Qiaokou District described a short-drama industry strategy that explicitly lists digital humans and “one-click storyboarding” among the technology advantages intended to become core production capacity inside a film-base cluster, indicating that regional media-industry development plans are treating digital-human assets as standardized production infrastructure alongside tooling for rapid audiovisual generation.
January 24 News
Entertainment: Mofa Technology (魔珐科技) was described as pushing AI digital humans into large-screen deployments across multiple commercial scenarios, positioning AI digital-human hosts as a next-generation human–machine interaction entry point and contrasting earlier “flat” animated overlays with newer host-style digital humans framed as having both “spirit” and “body,” implying higher-fidelity presence, richer performance control, and more lifelike hosting capabilities intended for real-world, venue-scale presentation and ongoing commercial use.
Marketing: Baidu (百度) was reported to have launched the Wenxin large-model 5.0 official release alongside “real-time interactive digital human” capabilities that stream and coordinate text, voice, and video in real time and integrate with digital-human motion libraries, with a high-profile demonstration involving a digital-human version of Luo Yonghao used for on-stage dialogue and referenced as applicable to content production and e-commerce livestreaming; Xinhua’s e-commerce coverage treated “anthropomorphic interactive services” as spanning digital-human livestream hosts, virtual customer service personas, intelligent shopping guides, and “AI companion-style” marketing conversations; JD.com (京东) was cited as using digital-human livestreaming at scale during Singles’ Day, with platform-wide GMV attributed to digital-human livestreams and large numbers of merchants participating, and digital-human livestreaming also positioned alongside shopping-assistant style applications as part of broader AI consumption behavior.
Education: Beijing Mengke Technology (北京萌科科技有限公司) was identified as the developer of GKK, presented as a workflow where uploading PPT/PDF materials can generate an interactive digital-human lecturer that delivers continuous instruction, can be interrupted, and answers questions in real time, with the operational promise that teachers or institutions can create many “digital clones” to provide individualized 1-to-1 explanations to very large numbers of students while keeping teaching logic and accumulated datasets within an organization’s own GKK knowledge base rather than depending on scarce star instructors; Guangdong’s practice-teaching coverage also described a classroom-like use of digital humans for real-time explanation and interaction within immersive XR/CAVE environments to support scenario-based learning experiences.
Healthcare: Hunan University of Chinese Medicine First Affiliated Hospital (湖南中医药大学第一附属医院) was reported to have launched the “Xintongzi” AI clinic-assistant as a mini-program fronted by a digital-human persona and powered by the DeepSeek model (DeepSeek 深度求索), enabling real-time dialogue for patient navigation and allowing configuration of the assistant’s displayed name and welcome message to function as a personalized intake and guidance aide during care-seeking.
Enterprise And Public Services: Qingbo Intelligence (清博智能) was described as an AI innovation service provider applying large-scale data and multimodal processing to domains including digital humans, large models, agents, AIGC creation, and AI short dramas, in the context of an industrial-base project landing in Ezhou; Yiyang (益阳) was reported to be developing a culture-and-tourism digital-human platform intended as a one-stop presentation layer for key city tourism content and services; Hangzhou Nanxingli (杭州南星里) was reported to have deployed the “Nanxing Zhilin” AI digital-human system (南星智邻) as part of a community service hub, with the digital human framed as an information and service interface embedded in neighborhood-scale operations; multiple local employment-service reports also described digital humans being used in job-matching and “livestream recruiting” workflows, including “digital-human plus human” dual-track livestreaming, enterprise-visit style livestreaming for remote “cloud touring,” and station-level experimentation with digital-human-supported outreach designed to expand channels and improve matching efficiency.
Governance And Ethical Contexts: China’s Ministry of Industry and Information Technology (工业和信息化部) was reported to be soliciting public input on a mandatory national standard project concerning metaverse classification and digital-human identity labeling requirements, explicitly aiming to standardize how digital-human identity information is recorded and how responsibility is attributed, with an approach summarized as assigning commercial and distribution-use digital humans a unique, code-based identifier to support authentication and accountability; separate legal-interpretation coverage framed “reviving” deceased individuals via AI and producing another person’s digital-virtual-human likeness or voice as raising infringement risks, including findings that natural-person voice rights may extend to AI-generated voices and that producing or distributing content using another person’s likeness in a digital-virtual-human form can implicate portrait-right violations, especially where the likeness is spoofed, mocked, or degraded.
January 23 News
Government And Public Services: Several China-based deployments frame virtual beings as front-end service agents or hosts inside public-sector workflows. In Changsha, a creative cartoon-expert digital human named 社社科科 was used to interpret and explain local social-science achievements in a public exchange activity format, positioning a branded synthetic character as an on-stage explainer for policy-adjacent content rather than as a consumer-facing influencer. In Chongqing, the municipal human-resources and social-security system was described as having built a “145N” digital human–enabled service体系 with full core-business digitalization, using virtual service agents as part of an integrated “digital HRSS” service architecture rather than a standalone avatar demo. In Ili (Xinjiang), the “我的伊犁” app’s government version was described as adding a government-service digital human called 小伊 alongside an AI-assisted 12345 hotline support capability, placing an avatar-like agent directly into citizen-facing intake and navigation. In Shanxi, a quality-infrastructure “one-stop” online platform was described as introducing a digital human called 小质, using a named synthetic persona as a service interface for standards/quality support. In Henan, an industry-and-commerce federation meeting’s “excellent work project” showcase was described as using an AI digital human host for the first time, applying a virtual presenter model to formal institutional events. In Hebei, Hebei IPTV’s ecosystem was described with a named assistant (河北IPTV灵犀助手) recognized as a “smart home new scenario” case and a separate digital human named 吉祥 recognized in a broadcast-regulator context tied to “digital stage” and “smart museum/heritage audiovisual systems,” treating virtual presenters and assistants as standardized components of home-media and cultural-display service stacks rather than novelty characters.
Enterprise And Commerce: Multiple reports position virtual beings as production tools for sales, customer service, and operational scale in China-based commercial settings, with emphasis on measurable efficiency and standardized deployment. JD.com (京东) was described as seeing AI-related search interest surge and as highlighting digital-human livestreaming as one of the application areas, framing the virtual host as a conversion-oriented retail interface tied to its supply-chain and C2M narrative rather than as a standalone media product. Baidu (百度) was described as experiencing a rapid expansion of “digital human” activity in which digital-human anchor orders were produced at high efficiency and were reported as outperforming human anchors in order-generation efficiency, with some livestream metrics described as leading, positioning the synthetic host as an industrialized performance unit for livestream commerce. Silicon Intelligence (硅基智能) was described as having output a very large number of “digital human agents,” framing its productization around agent-like digital humans delivered at scale, while other Nanjing-area references tied industrial software modernization to open-source operating-system ecosystems without detailing a specific named avatar. Zhejiang Wenhu Interact (浙文互联) was referenced in market coverage that attributed increased “consumption” or usage of a named digital-human product line called 派智, implying a metered or utilization-based deployment model where digital humans are consumed as a service rather than produced as one-off characters. Separate commerce-adjacent training narratives referenced “digital human anchors” in livestream e-commerce upskilling for merchants and practitioners, treating the avatar host as a reusable sales asset that can be taught and operationalized through training programs rather than only commissioned as custom animation.
Culture And Tourism: China-based cultural and tourism scenarios repeatedly anchor virtual beings in place-specific identity design and hybrid on-site media technologies, emphasizing localized styling and real-time performance. Shiyou Technology (Beijing) Co., Ltd. (世优（北京）科技股份有限公司) was repeatedly associated with a “digital human +” pattern in which city- or venue-branded characters are designed with local cultural signifiers and deployed across ceremonies, festivals, and visitor experiences; the Dali city digital human 小金花 was described as integrating Bai cultural elements and serving as a virtual host at the Dali Ethnic Film Week opening, driven by motion-capture for real-time interaction including Q&A, with attention to facial-expression control and natural body movement as the differentiator for on-stage credibility. The same pattern was also described as extending to other place-branded characters and staged city activations that combine a customized digital persona with AR overlays and glasses-free 3D or landmark projection-style presentation, treating the virtual being as the narrative “carrier” that binds multiple display technologies into a coherent tourism story. Leyard Lifu Culture (利亚德励丰文化) was described in a tourism-association context as presenting “digital human performances” together with “AI virtual restoration” for intangible-cultural-heritage activation and industry linkage, positioning synthetic performers and restoration pipelines as complementary tools for cultural resource revitalization rather than as entertainment-only novelties. In Hangzhou, a craft museum context described taking existing animated IP characters (御剑、婳伞、霓扇、匠宝) and producing 3D animation aligned to each museum’s cultural theme so the characters function as “digital human” style presenters tied to exhibition narratives, illustrating a pragmatic pathway where established IP is re-authored into venue-specific synthetic presenters instead of creating entirely new avatars.
Marketing And Brand Communication: A recurring China-based marketing logic treats virtual beings as controlled spokes-characters that can be reused across campaigns, events, and cross-IP collaborations while reducing scheduling and on-camera constraints. 37 Interactive Entertainment (三七互娱) was described as proposing the use of game IP linkage together with digital-human technology to empower National Games–related communication, positioning the digital persona as a brand-safe amplification layer that can bridge entertainment IP and large-scale civic-sports messaging. In Changsha’s local innovation ecosystem reporting, Wondershare Technology (万兴科技) was described as surfacing a named digital human application called 小麓, presenting a concrete example of a productized virtual being used to demonstrate AI capability within cultural scenarios and local digital-creative industry narratives, rather than as a purely speculative “AI+ culture” claim. Across these examples, the virtual being is framed less as an autonomous companion and more as an assetized media unit whose identity, voice, and appearance are deliberately bounded for repeatable institutional or brand communication.
Events, Media Personalities, And Synthetic Talk Formats: Virtual beings were also framed as substitutes or counterparts to recognizable media figures and as participants in staged dialogue formats, emphasizing performance evaluation and audience-facing credibility. A report describing a “Moment大会” demonstration referenced a “Luo Yonghao digital human” being presented in a live event setting and then evaluated by a media person with a numeric score alongside an internal self-assessment, illustrating an emerging evaluative culture where a digital replica or synthetic persona is judged on conversational performance and perceived fidelity to a known public figure rather than on pure animation quality. In parallel, multiple institutional-event references to “AI digital human hosting” show that the synthetic presenter format is being normalized for formal agendas, with the host role treated as a deployable function that can be swapped in without changing the underlying event structure, thereby turning the “virtual host” into a standardized media role rather than a one-off spectacle.
Transportation And On-Site Service Terminals: Some cases explicitly connect digital humans to physical service touchpoints without framing them as robots, using the avatar as the face of an on-site “one-machine” service workflow. A Nanning airport scenario referenced a named digital human 桂飞飞 described as successfully launched in the context of smart-airport construction, presenting the virtual being as a milestone interface for passenger guidance and service experience modernization. In another deployment, Chongqing Meituan San快 Microcredit Co., Ltd. (重庆美团三快小额贷款有限公司) was described as using an AI customer-service digital-human integrated terminal (数字人一体机) to upgrade “smart travel” service experiences, framing the digital human as a kiosk- or counter-facing service persona integrated with back-end customer-service logic, rather than as a purely online livestream character.
January 22 News
Healthcare: Hangzhou established a youth-oriented AI companion digital human called Le Xiaoming (乐小明), developed under the leadership of Hangzhou Municipal Health Commission with participation from Professor Luo Hong of the Zhejiang University School of Medicine Affiliated Mental Health Center (Hangzhou Seventh People’s Hospital), positioning the system as an adolescent psychological counseling “intelligent agent” digital human intended for conversational mental-health support and structured guidance. Ant Group (蚂蚁集团) extended its Ant Aifu (蚂蚁阿福) medical service by opening “AI avatar/clone” capability to physicians, with more than 1,000 doctors creating their own AI “copies” on the platform to provide 24/7 responses to users’ health consultations, and the PC-side upgrade adding a DeepSearch function aimed at improving clinician-facing information retrieval and consultation handling.
Marketing: JD.com (京东) partnered with Red Magic (红魔) to run an accelerated device upgrade campaign tied to early ordering for the Red Magic 11 Air, and used a digital-human livestream “debut” as the public-facing activation format; the deployment was described as being deeply integrated with JoyAI (JoyAI), framing the virtual character not as a one-off mascot appearance but as an AI-enabled livestreaming host for commerce conversion and campaign execution. Tiandi Online (天地在线) was described as operating in “virtual digital human + virtual IP + digital collectibles” linkages, including creation and operation of digital personas and services for “digital idols,” and publishing a proprietary virtual IP named Yuanqi (元启) alongside a co-creation relationship with UME Cinemas (UME影城) positioned around metaverse-linked content activity.
Real Estate And Legal Services: Zhengzhou courts introduced an “安心购” auction pathway that paired enforcement actions with AI digital-human livestreaming for “cloud house-viewing,” describing the digital human as a property guide that presents information in a transparent, comprehensive way during online viewing, while enforcement personnel address post-sale handover obstacles to reduce buyer risk; the case is explicitly framed as a court-linked workflow combining virtual guided presentation with offline enforcement to support judicial auction transactions.
Enterprise: Kingsoft Office (金山办公) presented WPS 365 (WPS 365) as a workplace system where “digital employees” function as the primary AI office-interaction interface, including a mechanism to create a departing employee’s digital avatar for knowledge handover, and a digital-employee development platform for enterprises to build customized AI tools. Nanjing Side Exhibition Technology (南京思德展示科技股份有限公司) disclosed a patent application for an exhibition-hall intelligent report-generation system that fuses multimodal data into a unified semantic space, explicitly including “digital human interaction” among the heterogeneous data sources to map behavior signals to inferred interests and intentions for reporting and decision support.
Public Services: Suzhou Housing Provident Fund Administration (苏州市住房公积金管理中心) operationalized an AI digital-human customer-service role and a product named Su Xiaojin Cloud Classroom (苏小金云学堂), alongside an AI service zone in the service hall and 5G-enabled remote service, explicitly describing the service-mode shift toward side-by-side assistance supported by digital-human and AI tooling. Shiyou Technology (世优科技) was associated with an AI digital human named Xiya (希娅) included in a metaverse typical-case list issued by multiple national ministries, and its “digital human +” application approach was described as spanning government-service halls and corporate exhibition environments; a separate public-service example referenced an airport digital human named Gui Feifei (桂飞飞) at Nanning Airport that answers operational queries such as flight status and lost-and-found, positioning the character as a front-line service agent for passenger information and routine service triage.
Culture And Education: Nippon.com (日本国际交流基金会相关媒体平台) described an education-channel case called Free and Easy V (自由自在V) built through industry–academia collaboration, hosted by a folklore-themed virtual broadcaster Morohoshi Meguru (诸星Meguru) and a “zombie teacher” character, using virtual-host formats for education-oriented programming. A Beijing University training-class setting referenced “digital human coach” instruction, indicating digital-human IP and AI coursework contexts where virtual-being production and operation are treated as professional skills within new-media and IP-oriented curricula.
XR, Rendering, And Production Technologies: Hanbo Semiconductor (瀚博半导体（上海）股份有限公司) filed a patent application for real-time digital-human rendering methods and devices designed to handle high frame rate, low latency, and scalable rendering requirements, treating digital-human rendering as a performance-critical pipeline problem. ArcSoft (虹软科技) described AI-XR work that includes virtual digital-human expression driving and body driving for interaction, alongside visual-presentation technologies such as video see-through (VST) and asynchronous time/space warp methods intended for XR performance and perceptual stability, positioning digital-human animation control and XR display compensation as an integrated stack. Qingtong Vision (青瞳视觉) was described as providing motion-capture technology and end-to-end technical support for virtual digital-human live production, including a named case supporting Migu NBA (咪咕NBA) playoff/finals virtual livestreaming, framing optical motion capture as the enabling layer for broadcast-grade virtual-character performance and event coverage.
January 21 News
Education: In Chongqing’s Dazu District, a joint “graduation” showcase for four senior-teacher studios included a case where a trainee representative delivered a teaching-research outcomes briefing in the form of a digital human presentation, indicating direct use of a human-like virtual presenter as a reporting interface in a teacher professional-development context. In Jiangxi, the provincial “AI+” action plan explicitly calls for accelerating standardized smart-campus建设 and for embedding AI and digital-human technologies across the full set of teaching and learning elements and processes, alongside encouragement for capable primary and secondary schools to build an “AI sports bar” style campus application space; this positions digital humans as a planned, system-level instructional and campus-service modality rather than a one-off pilot. In Hefei, Hongqiao Primary School’s AI classroom practice is described as using digital-human scenario simulation and an interactive “venting/talking” segment to support students’ emotion recognition and regulation skills, making the digital human a structured component of socio-emotional learning activities rather than only a content narrator.
Government And Public Services: In Haikou, the Haikou Maritime Safety Administration is described as deploying an intelligent voice digital human named “Xiao Haishi” (小海狮) in conjunction with a ship-registration intelligent form-filling system, with the stated operational goal of improving data-entry and service handling through a conversational, personified interface in passenger-transport safety and maritime administrative services. In Henan’s Hebi, a conscription publicity livestream on short-video platforms used a hyper-realistic 3D virtual digital human called “Rong Zhiguang” (戎之光) as a dedicated national-defense education and recruitment promotion persona, and the persona is attributed to in-school development by teachers and students from a “Lijian Propaganda Team” (利剑宣传队), making it a concrete example of a locally built synthetic character used in public messaging and live online outreach.
Marketing And Commerce: Zhejiang Wenhuan (浙文互联) is reported as collaborating with Ocean Engine (巨量引擎), a ByteDance (字节跳动) advertising business, to co-build “Paizhi” (派智) agile/animated digital humans, with the context framing this as AI marketing-oriented deployment and linking performance claims to rapid growth in usage/consumption metrics; this is a concrete example of a platformized digital-human product positioned for scaled advertising and brand communication workflows. Tiandi Online (天地在线) is described as having built an AI marketing “full-chain” capability that explicitly includes digital humans, XR livestreaming, digital content production, and digital scene construction, framing digital humans as one component in an integrated pipeline for campaign production and live commerce presentation rather than a standalone avatar tool.
Enterprise Productivity And Digital Employees: WPS 365 (WPS 365) is described as providing mechanisms for creating a departing employee’s “digital double” for knowledge handover, and as offering a “digital employee” development platform that lets enterprises customize AI tools; this treats the virtual-being construct as an internal-facing, role-based agent/persona anchored to organizational knowledge transfer and configurable workplace automation rather than an external brand avatar. Separately, a “product-level open-source digital double OS” is described as enabling plug-and-play “skill packages,” framed around enterprise AI tool fragmentation and the goal of making AI operate more like a coherent human employee, which positions the “digital double” as an extensible operating layer for assembling capabilities into a consistent worker-like interface.
Tooling And Production Technology: Mofa Technology (魔珐科技) and Shanghai Mowu Technology (上海墨舞科技) are reported as filing a patent application for methods that generate digital-human video素材 and digital-human video more generally, emphasizing reducing the user’s operational threshold; this is a concrete production-side case that treats digital humans as generated media assets and focuses on workflow simplification for creating reusable video components. V2Fun.art (V2Fun.art) is described as a one-click AI character generation tool positioned to help users create “digital characters” for game roles, virtual idols, and advertising promotional figures, framing the primary use case as lowering technical barriers to producing synthetic characters for downstream deployment in entertainment and marketing content pipelines.
Culture, Entertainment, And Virtual Idols: A virtual persona “Cai Caizi” (菜菜子), described as a “clone”/virtual image associated with performer Cai Ming (蔡明), is presented as a cross-generational virtual idol experiment and an example of turning a well-known human performer into an operational virtual idol identity with ongoing ranking/visibility outcomes. Separately, the virtual streamer Aza (阿萨Aza) is referenced in the context of sustained prominence in virtual broadcasting, used to illustrate the competitiveness and maturation of the virtual-host ecosystem as AI capabilities increase the functional expectations for virtual personas. Alongside these cases, AI companions are explicitly distinguished as systems oriented toward personalized emotional dialogue and companionship, situating “companion-like” synthetic characters as a recognizable category adjacent to performance- and hosting-centric virtual idols.
Tourism And Cultural Heritage: In Zhejiang’s Ningbo, a listed digital-culture-and-tourism case at Minghe Ancient Town (鸣鹤古镇) includes a 3D digital human interaction component paired with interactive experiences such as a “Minghe pinball game” (鸣鹤弹珠游戏) and a “medicine-making machine” (良药制造机) smart device, which is a concrete example of embedding a human-like interactive avatar into an on-site cultural heritage visitor experience to support engagement and scenario-based interpretation rather than relying only on static displays.
Compliance And Governance: In Jiangsu, policy measures for “AI OPC” development include instituting an enterprise contact-point mechanism specifically for digital-human livestreaming OPC firms and drafting and issuing a digital-human livestream compliance guideline aimed at helping enterprises mitigate marketing risks; this is a concrete governance case that treats digital-human livestreaming as a distinct operational category requiring risk controls, compliance expectations, and dedicated administrative support structures.
January 20 News
Finance: Chongqing Bank (重庆银行) described building a “personal mortgage loan digital-intelligence employee” agent based on AI, positioned as a digital-human-style frontline worker for retail mortgage processing, with an explicit case claim of “seconds-level approval” for high-score customers and an accompanying note that a “digital human system” optimization-and-renodeling procurement project for 2025 reached a winning-bid announcement stage, implying the bank is treating the digital-human stack as an upgradable enterprise system rather than a one-off demo; Jiufang Intelligent Investment (九方智投) was referenced as offering a “Jiufang Digital Human” capability alongside intelligent image recognition, AIGC content creation, sentiment/tendency judgment, public-opinion analysis, and financial knowledge services, framing a finance-oriented digital-human front end as part of a broader AI toolchain for content, analysis, and advisory-style interaction rather than a standalone avatar.
Marketing: United Creation Technology (贵州联合创世科技有限公司) was reported to have launched a fourth-generation digital-human live-streaming system in 2025, explicitly tied to “AI customer acquisition” for live commerce and multi-industry use, treating the digital human as a sales and lead-generation performer that can be deployed across verticals; Keyi Cloud (客易云) was cited in connection with “TikTok digital humans,” positioning synthetic presenters as a foreign-trade growth tool on TikTok for outbound customer development, with the platform named as the distribution channel rather than the creator; Gree Electric Appliances (格力) was referenced around the opening of the “Dong Mingzhu Health Home” Northwest flagship store in Xi’an, with the digital-human mention appearing as part of the retail experience/marketing ecosystem context in the same cluster of reporting rather than as a fully specified in-store avatar implementation.
Enterprise: Kaipu Cloud (开普云) was described as applying speech-recognition technology into a “digital human interaction platform,” indicating a concrete architecture where ASR is a core enabling layer for interactive digital-human experiences; Guangdong Southern Media (广东南方新媒体股份有限公司) was mentioned as using technologies including digital humans, implying a media-technology operator integrating synthetic presenters or digital staff into content production and/or service workflows as a named applied technology area rather than an abstract research topic.
Government And Public Services: Xuzhou City Operation Comprehensive Supervision Platform (徐州市城市运行综合监管平台) was reported to include a digital human named “Xiaozhi” (小智) that can conduct dialogue-style interaction and rapidly retrieve and present operational information for “city lifeline” supervision, presenting a concrete “digital staff” use case embedded in municipal command-and-control; Jiangsu provincial policy coverage described establishing a “digital-human live-streaming OPC enterprise contact-point system” and issuing a “digital-human live-streaming compliance guideline,” treating synthetic hosts for live commerce as a regulated operational category with institutional touchpoints rather than purely a market behavior; Putuo District of Shanghai (上海市普陀区) was reported to have introduced a virtual digital human IP “Pu Xiaotuo” (普小陀) as an officially designated “digital lecturer,” replacing a conventional human host for an immersive youth-theory presentation format and explicitly institutionalizing the avatar as a role-bearing presenter.
Culture And Education: Sichuan Radio and Television (四川广播电视台) was described as co-developing the “Zaxika’s Meiduo” (《扎溪卡的梅朵》) project as a children’s picture-book-centered cultural assistance initiative that also links an AI digital human “Meiduo” (梅朵) to science-popularization cultural-creative product development, presenting a concrete “character IP + digital human” extension pattern; Kunming Information Port (昆明信息港) described intangible cultural heritage digitization where dancers’ motions are captured into precise skeletal data and then used to generate an interactive, teachable digital-human model, specifying the technical pathway from performance capture to a pedagogical avatar; Changtai District cultural-heritage reporting framed “digital doubles” (数字分身) for historic architecture as a structured “digital archive” combining 3D models with rubbings, survey drawings, and multi-angle historical photos, showing a specific archival use pattern where the “digital body” is the access interface for preservation and interpretation; Zhengzhou Blind-Deaf-Mute School (郑州盲聋哑学校) was reported to use AI digital humans in classroom practice, including a “digital human crossing into the classroom” scenario within Chinese-language teaching and broader examples of using AI to efficiently produce courseware, presenting an applied special-education case in which synthetic characters support instruction rather than entertainment.
Healthcare: Hangzhou municipal health reporting described an occupational-health AI digital human “Zhifang Yunpeng” (职防云鹏) introduced within the city’s health system to popularize workplace health and occupational-disease prevention knowledge in a more vivid and accessible format, positioning the digital human as a public-health explainer embedded in official health education channels rather than as a consumer chatbot.
Entertainment: Bilibili (哔哩哔哩) reporting on annual creator honors referenced a celebrity “digital double” style virtual persona “Caicaizi” (菜菜子) associated with Cai Ming (蔡明) and a virtual musician “Asa Aza” (阿萨Aza) receiving major platform recognition, with the same cluster of coverage noting established virtual UP creators such as “Jia Ran” (嘉然) in the competitive ecosystem narrative, collectively presenting concrete examples of synthetic-character performers treated as first-class creators inside a mainstream video platform’s award and ranking system.
January 19 News
Government Services: Xinjiang Human Resources and Social Security Department (新疆维吾尔自治区人力资源和社会保障厅) described building an integrated “Digital HRSS” framework under a unified “one network” initiative, positioning digital-human capabilities within a broader provincial-level platformization effort for HRSS services, even though the publicly surfaced description emphasized infrastructure and system integration rather than a specific named digital-human instance, persona, or deployment scenario.
Entertainment: Changchun Film Studio Group (长影集团) was reported to have award-recognized works that applied an “AI historical scene restoration system” together with digital-human techniques to reconstruct historical narrative elements connected to the songwriter Cao Huoxing and his family story, framing digital humans as a production tool for recreating people and scenes with documentary-like specificity; separately, Alibaba Digital Media and Entertainment (阿里大文娱) was reported to have generated a hyper-realistic virtual digital actor named Lili (厘里) via in-house AI, emphasizing fine-grained facial micro-expressions, facial muscle motion, and eye-gaze/emotion conveyance as the core realism stack, and it was also linked to the Hangzhou Asian Games opening sequence in which 105,791,208 global “digital torchbearers” were aggregated into a single figurative outcome, functioning as a mass-participation avatarization case where individual digital identities were composited into a unified on-screen virtual form.
Marketing & E-commerce: Guangdong Geek Marketing Technology Co., Ltd. (广东极客营销科技有限公司) appeared in cross-border trade/e-commerce evaluations as providing an “AI digital employee” and end-to-end customer-acquisition solutions that explicitly included a digital-human short-video matrix system and an AI digital-human livestream product (product name: 极客AI直播机器人) used in at least one cited 3C-category scenario as part of a combined acquisition-and-ranking workflow; Semir Group (森马集团) was described in an e-commerce visual-marketing context as using virtual digital-human livestreaming with rapid scene switching (for example, seasonally appropriate background changes) to increase immersion, alongside a broader shift from flat images to 3D and video-based presentations aimed at “what you see is what you get” product visualization.
Cultural Heritage & Tourism: Xixian County publicity authorities (隰县宣传系统) described a heritage-IP activation approach around “Little Western Heaven” (小西天) that included a digital-human dialogue experience enabling visitors to “converse” across nearly 400 years, presenting a case-study pattern in which a historically situated persona is instantiated as an interactive digital character to support cultural storytelling and local heritage branding.
Enterprise & Collaboration: BizConf Telecom (会畅科技) was reported as expanding an “AI application” concept footprint tied to metaverse-adjacent workflows, explicitly naming 3D cloud video scenarios that include 3D model building, rendering, and rapid generation of AI virtual digital humans, positioning virtual-human creation as a native capability inside enterprise-grade 3D communication and content pipelines rather than a standalone marketing asset.
Consumer Companions: Fangzhi Technology (方直科技) stated that it launched the “Yangzai AI Companion Device” (阳仔AI陪伴机) in mid-October 2025 and had already placed it through JD.com self-operated e-commerce channels (京东自营) and offline distribution, making it a concrete AI-companion deployment case framed as a productized companion application rather than a generic chatbot, and explicitly associated with the broader “virtual digital human” concept set referenced in its market disclosures.
January 18 News
Cultural Heritage: A museum interpretation program in Shaanxi opened a dedicated digital-human docent hall, positioning “数字人” as a structured exhibit-guidance interface for on-site visitors within the “Cultural Relics Are Speaking” training initiative that brought together young Chinese and international museum explainers from Tsinghua University (清华大学) and Shaanxi University of Chinese Medicine (陕西中医药大学) in Xianyang, using the digital docent setting as a practical environment for interpretive narration and visitor engagement around historical collections.
Government And Public Services: Multiple local-government and public-service deployments framed digital humans as front-end service actors in administrative and civic workflows, including an AI virtual digital-human host used in Mianyang’s public resource transaction system to support a “full-process” electronic trading experience, and the New Beiqiu District “Xiaoxin” (小新) digital human delivered across three endpoints to connect high-frequency government services spanning market access, project construction, talent employment, and public services, with an associated “AI + environmental impact assessment” pre-review capability that used graded warning cues to assist early-stage review speed and consistency.
Enterprise And Operations: China Telecom Jiangsu (中国电信江苏公司) was presented in a municipal strategic-cooperation context through a dedicated telecom digital-human introduction tied to the “Zhiyun Taizhou” (智云泰州) and “Zhiyun Store Taizhou Zone” (智云商店泰州专区) concept, framing the digital human as a demonstrable interface for enterprise-and-government customers to understand packaged cloud-service capabilities and localized solution delivery through a storefront-style service aggregation model.
Entertainment And Live Production: A large-scale youth gala production in Nanning integrated digital humans and virtual avatar-frame products as on-screen creative elements to translate festive visual motifs into “screen-first” experiences, while Lenovo (联想) described a sports-broadcast style use of virtual character imagery rendered as 3D animation to support referee-decision visualization—specifically for offside replay explanation—positioning virtual figures as clarity-enhancing presenters for live and online audiences within a match-analysis workflow rather than as narrative characters.
Marketing And Digital Commerce: Zhejiang Wenhu Interact (浙文互联) was described as integrating DeepSeek (深度求索) capabilities—explicitly including multimodal generation and natural-language processing—into its digital marketing and virtual-human business scenarios, framing the virtual-human layer as a production-facing application surface for campaign operations and content workflows powered by large-model and multimodal toolchains rather than a single character deployment.
Patents And Core Production Technologies: Baidu Online Network Technology (Beijing) Co., Ltd. (北京百度网讯科技有限公司) disclosed a patent application for a digital-human interaction method that generates response videos for interacting with a target entity, specifying a pipeline in which the output is an “answer video” suitable for video-based interactive exchanges; Hisense Visual Technology Co., Ltd. (海信视像) disclosed a patent application focused on improving digital-human animation efficiency by extracting sports/motion terms, mapping them to animation-control parameters, and using those parameters to drive the digital-human animation process as a structured control workflow; iFLYTEK Open Platform (讯飞开放平台) described an end-to-end capability set for “超拟人” interaction that explicitly included long multi-turn dialogue, AI persona, empathy-style interaction, long-term memory, and “driving digital humans,” emphasizing a low-latency full-chain response profile as a platform-level production and runtime foundation for digital-human experiences.
Enterprise AI Systems And Solution Stacks: Beijing Xinghe Zhuoyue Technology Co., Ltd. (北京星河卓越科技有限公司) described its focus on building long-running AI system capabilities for government-and-enterprise customers, explicitly listing digital humans and multimodal generation alongside content generation/review and privacy-oriented solution components, positioning the digital-human layer as one module within an integrated enterprise AI stack rather than a standalone character product; Deep Cloud Sea Intelligent Technology (Shanghai) Co., Ltd. (深度云海智能科技(上海)有限公司) and Shaanxi Bode (陕西博德) convened an AI search technology and application innovation summit in Xi’an that explicitly referenced digital humans in the conference context, framing them as part of a broader application landscape connected to search-oriented AI workflows and deployment discussions.
Media And Content Production Platforms: Shandong Shuwen Group (山东数文集团) presented a product cluster that included a “media full-type digital-human platform” branded Yidian Tiancheng (壹点天成), alongside a media large-model offering Yidian Tianji (壹点天玑), a multimodal content-risk-control platform Yidian Tianyan (壹点天眼), and a metaverse activity platform Yidian Tianyuan (壹点天元), collectively framing digital humans as one of several interoperable content-production and governance capabilities within a media-industry model-and-platform suite.
Regulation And Risk Controls: A live-ecommerce regulatory update explicitly stated that digital-human streamers would be brought under live-streaming e-commerce supervision, and a related risk discussion highlighted two concrete failure modes—identity forgery through highly realistic virtual personas and prompt-induced large-model outputs that can violate legal or ethical constraints—framing the operational requirement as “using AI to govern AI” through traceability and enforcement mechanisms aimed at virtual-persona authenticity and controllable generative behavior; in workforce and public-employment outreach, a “real person + AI digital human” livestream format was used for recruitment activity delivery, with a government employment team running broadcasts via a Douyin account, illustrating how policy-facing communication formats are already blending human presenters with digital-human components in regulated public-information settings.
January 17 News
Healthcare: Nanjing University (南京大学) reported the launch in Chongqing of a focused-ultrasound “medical quality control system + digital platform” initiative that included the release of an “AI digital human” innovation persona used as a concrete public-facing interface for introducing the platform and its functions in a medical quality-control and discipline-development context. Alibaba Health (阿里健康) was described in investor commentary as benefiting from the popularity of “Ant Afu” as an AI “digital avatar/doppelgänger” health service that goes beyond basic chat by integrating health-check report interpretation, AI-assisted consultation, and chronic-disease management workflows, framed as a productized, high-frequency health-advice interface. Separate consumer-health reporting on sleep monitoring described a linkage pattern in which abnormal home sleep data can be routed to a top-tier hospital “AI avatar/doppelgänger” service associated with an academic medical team, positioning the avatar as a triage-and-advice layer that escalates by risk level rather than as a standalone wellness chatbot.
Education: Tianjin University of Science and Technology (天津科技大学) was reported as digitizing student-aid and certification procedures through online workflows, while Tianjin Electronic Information Vocational and Technical College (天津电子信息职业技术学院) deployed two always-available AI digital-human counselors, “Xiaode” and “Xiaoyi,” with 24-hour voice interaction positioned as a student-facing support role. Nankai University (南开大学) reported a course-oriented “virtual teaching and research office” that built digital-human models of four instructors and produced 48 instructional videos covering difficult and key points, treating teacher digital humans as a scalable delivery mechanism for structured micro-lessons rather than as entertainment avatars. A separate school activity report from Zhengzhou’s Economic and Technological Development Zone referenced classroom use of multiple AI tools including “digital humans,” indicating that digital-human interfaces are being treated as one of several practical, named tool categories integrated into media literacy or AI-supported teaching demonstrations rather than as a single monolithic system.
Government & Public Services: Changzhou Xinbei District Government (常州市新北区人民政府) described a “Xiaoxin” digital human deployed across three client endpoints and connected to 820 high-frequency government service scenarios spanning market access, project construction, talent employment, and livelihood services, presented as a unified front-end persona over a large catalog of transactional and advisory workflows; the same report also referenced an “AI + environmental impact assessment” intelligent pre-review that uses color-coded warnings to support faster preliminary review, showing the digital-human layer operating alongside automated decision-support. Mianyang Municipal Government (绵阳市人民政府) described an AI virtual digital-human host used inside a public-resources trading service center’s digitized bidding environment, including a first appearance in a mining-rights auction where the virtual host worked in tandem with an on-site auctioneer, making the digital human a ceremonial-and-procedural interface for high-stakes public transactions rather than a back-office automation. Employment-service reporting from Linyi framed “human + AI digital human” around-the-clock livestream job-matching as an operational model for labor-market services, indicating that digital humans are being used as persistent on-camera presenters to sustain continuous service windows that human staff alone cannot cover.
Finance & Commerce: Guotai Haitong (国泰海通) was described as building a “digital human avatar/doppelgänger system” that turns employees and investment-advisory roles into digitized hosts, with automation that captures hot topics, generates livestream scripts, and outputs a short, distributable video in roughly five minutes in a “host–guest” dialogue format, using the digital humans as the visible presenters of market commentary rather than purely internal summarization tools. A separate cross-border e-commerce case-study write-up described Customer Easy Cloud (客易云) as using digital-human livestreaming to build trust for international audiences, including an example in which a maternal-and-infant brand streamed real-time factory footage and invited a third-party testing organization to verify ingredients during the session, positioning the digital human as the continuity anchor for a credibility-focused, evidence-backed broadcast format rather than as a novelty presenter.
Marketing & Livestream Commerce: Baidu (百度) was repeatedly positioned around digital-human livestreaming as a productized commercialization pathway: coverage of its “Huiboxing” offering described a “Luo Yonghao digital human” as a flagship “super head” digital-host implementation and cited a county-level example in Bijie, Guizhou where a deer-antler livestream room reportedly achieved a threefold sales increase using Baidu’s digital-human livestream capability, illustrating the model of pairing a recognizable persona or a local specialty product with a standardized digital-presenter stack. A separate speech summary attributed to Baidu executive Ping Xiaoli framed “virtual humans/digital humans” as moving through staged capability generations toward a “4.0” phase, with earlier phases characterized as having visible limitations in facial expression and voice naturalness, implicitly aligning Baidu’s commercial roadmap with measurable gains in realism, interaction quality, and scenario coverage that underpin the livestream and e-commerce examples already in circulation.
Culture & Tourism: Sohu coverage of Nanjing’s inclusion in national metaverse “typical cases” referenced Silicon Intelligence (硅基智能) through a named case, “an AIGC digital human intelligent interactive platform enabling multi-industry metaverse interaction,” presented as an example of how a digital-human interaction layer is being packaged as reusable infrastructure across sectors rather than as a single venue-specific avatar deployment. FanTuo Digital Creation (凡拓数创) described museum and exhibition experiences in which a digital-human employee persona, “Ai Wenwen,” provides an “immersive” guided experience for cultural exhibitions, indicating a use pattern where a branded, staff-like synthetic character functions as the consistent interpreter and guide across rotating exhibitions and venue programs. Shanghai Putuo District (上海普陀区) was reported as running an “unhosted” immersive youth theory-presentation format in which a virtual digital human leads the audience through scenario-based enactments and multimedia linkage, using the digital human as the narrative spine that replaces a traditional stage host and coordinates transitions among mixed media elements.
Enterprise Platforms & Developer Ecosystems: iFlytek Open Platform (讯飞开放平台) was described as emphasizing “highly human-like” interaction in its large-model phase, explicitly listing capabilities such as ultra-long multi-turn dialogue, persona configuration, empathetic-style interaction, long-term memory, and digital-human driving, and quantifying end-to-end response time as part of the platform narrative; in parallel, education-industry reporting tied to its Spark education model described a “teacher–student–machine” triadic classroom architecture at large scale and cited “virtual human Q&A” usage volume, treating virtual humans as an interaction modality embedded in platformized education delivery rather than as standalone characters. A separate policy-text repost oriented to Sichuan’s digital-economy measures enumerated “digital human livestreaming” among the operational categories addressed in the wider digital-services and e-commerce ecosystem, signaling that digital humans are being treated as a recognized production format in regional economic policy toolkits rather than as a niche experiment.
Devices & Patents: Zhong’an Netvision Metaverse Technology (Xiamen) (中安网视野（厦门）元宇宙科技股份有限公司) together with Xiamen Yingcai School (厦门英才学校) disclosed a patent application for an “AI digital human intelligent photo frame knowledge Q&A system” with real-time question answering, representing a concrete device form factor in which a digital human is embedded into a household or classroom display object and positioned as an always-on conversational interface for knowledge retrieval and explanation rather than a phone-bound avatar. The repeated appearance of “digital human” in government-service halls, auction venues, and education media also implies a growing pattern of deploying digital-human front ends on large screens and kiosks, but the only explicitly specified consumer-device instantiation here is the photo-frame Q&A patent.
Ethical Contexts: Hangzhou-focused commentary on “digital garbage” framed low-quality AI video, incoherent AI novels, and on-demand AI virtual roles as part of a broader content-quality and attention-economy problem, explicitly naming “AI virtual characters” and “digital humans” as one of the forms through which mass-produced, low-signal synthetic content circulates; the core concrete implication is that the same rapid-generation pipelines that enable practical digital-human presenters in services and commerce are also being used to generate disposable character content at scale, creating a governance and quality-control challenge that sits adjacent to legitimate deployments rather than replacing them.
January 16 News
Cultural: In Nanchang, China Unicom Jiangxi (江西联通) deployed an AI digital human named Huánhuán (环环) at the Jiangxi Provincial Museum as an on-site interpretive interface tied to the museum’s collections, explicitly referencing the Xin’gan Shang-period tomb finds and using the excavated jade “feathered-human” artifact motif (玉羽人) as the digital human’s cultural anchor; the same local cluster framing also connected the provincial museum, library, and science museum in a riverside cultural zone, positioning the digital human as part of a broader “culture made speakable” digitalization approach that treats artifacts and public cultural resources as interactive, personified interfaces rather than static displays.
Entertainment: Chengdu Digital Sky Technology (成都数字天空科技有限公司) was described as building a “hyper-realistic digital human” production capability by accelerating an intermediate testing platform for ultra-realistic digital-human modeling and animated performance (超写实数字人建模与动画演绎中试平台), with the stated orientation toward mobile-end product R&D and distribution and toward producing Chengdu-origin digital cultural-creative “hits,” implying digital humans as core assets for entertainment-grade character performance and content packaging.
Media And Broadcasting: Guangdong Radio and Television (广东广播电视台) and Lianhui Technology (联汇科技) were reported as jointly implementing an “intelligent converged-media application” that integrates digital humans into the end-to-end content-production workflow, presenting the digital human as a reusable production component inside newsroom and broader media operations rather than a one-off on-camera novelty; separately, Shandong Radio and Television (山东广播电视台) was described as deploying ultra-realistic virtual digital humans named Hǎilán (海蓝) and Dàiqīng (岱青) across news presentation, e-commerce livestreaming, and entertainment interaction, including a sign-language digital human capability associated with Hǎilán, framing broadcast virtual beings as multi-scene presenters and interactive hosts spanning public communication and commercial conversion.
Marketing And Brand Activation: Aokang (奥康) was presented as iterating from an initial brand virtual person “Aoki” (Aoki) to an AI-enabled fashion show and broader “full-link content ecosystem,” treating the virtual person as both a digital identity carrier and a scalable content-production and campaign-orchestration node; in parallel, Tianyu Digital Technology (天娱数科) described ongoing work to refine cross-modal intelligent algorithms using large-scale data accumulated on its Yuanxiang Platform (元享平台), with explicit exploration of virtual humans as an application direction, positioning virtual beings as outputs of multimodal modeling pipelines rather than purely front-end spokespersons.
Commerce And Livestreaming: Lao Guanjia (老管家) was reported as expanding to nearly ten Kuaishou (快手) storefronts and turning to digital human livestreaming as a practical substitute for staffing multiple human hosts, using virtual beings to stabilize labor costs while maintaining continuous selling capacity across parallel shops; a separate case described a Baidu (百度) digital-human livestream that replicated a real person’s voice, facial expressions, and speaking style to a high degree and achieved reported high GMV in a six-hour session, aligning the use case with conversion-centric e-commerce hosting rather than entertainment-only performance; another commerce-oriented deployment framed a “digital human livestream system” as an export lead-generation tool on TikTok, where the system was credited with breaking language and time-zone barriers for foreign-trade outreach and with adding user-behavior tracking and modeling to support conversion analysis, illustrated through a Zhejiang toy manufacturer producing a “children’s toy safety guide” video series via the digital-human workflow.
Transportation And Travel Services: Shiyou Technology (世优科技) deployed an AI digital human named Guìfēifēi (桂飞飞) as a set of twelve intelligent customer-service terminals inside Nanning Airport’s T2 terminal, framing the virtual being as a standardized front-desk layer for passenger guidance and service delivery in a high-throughput public-transport environment where consistent responses and always-on availability are operationally valuable.
Education: Nanjing’s rollout of a city-level preschool “intelligent agent” called Níngxiǎozhì (宁小智) was described as having two digital-human IPs (双数字人IP), indicating a designed pair of characterized virtual beings used as the public-facing embodiment of an educational agent system for early-childhood contexts; separately, a Chongqing middle-school “science mini-classroom” was described as being delivered by a digital-human teacher, using the virtual being as the presenting instructor to guide students through everyday-science topics in short-form lessons.
Enterprise And Property Operations: Phantom Future (幻影未来) was reported as applying AI digital humans across multiple properties operated by SINO Group (信和集团) in Hong Kong, with named deployments including China Hong Kong City (中港城) and One Capital Place (朗壹中心) and additional rollouts noted at Tsim Sha Tsui Centre (尖沙咀中心) and Empire Centre (帝国中心), framing the virtual being as a tenant- and visitor-facing service layer embedded into commercial real-estate operations and scalable across a property portfolio.
Technology And R&D: Diejing Digital Technology Shanghai (叠境数字科技（上海）有限公司) was profiled through its CTO’s work spanning light-field technology, 3D reconstruction, and digital humans, positioning digital-human capability as part of a broader spatial-capture and volumetric-media toolchain; Hanbo Semiconductor (Shanghai) (瀚博半导体（上海）股份有限公司) disclosed a patent application covering methods for generating videos containing digital humans and methods for training digital-human generation models, explicitly treating the virtual being as a learned generative target within a video-synthesis pipeline rather than merely a rendered character asset.
Ethical Contexts And Abuse Cases: Shanghai police disclosed a fraud case in which the “woman” targeted in an online dating scheme was an AI virtual person, with the interaction gated behind paid “top-ups” to continue chatting and the reported loss scale exceeding ten million RMB, illustrating a concrete misuse pattern where synthetic personas are operationalized as persistent social-engineering fronts that simulate intimacy to drive repeated micro-transactions rather than one-time theft.
January 15 News
Fraud And Consumer Harm: Shanghai police disclosures and related reporting described a dating-app fraud pattern in which victims were induced to repeatedly top up in-app “virtual coins” to unlock text chats and video calls, while the “female” persona was an AI virtual being/AI-enhanced virtual avatar rather than a real person, illustrating how synthetic characters can be deployed as persuasive front-ends in relationship scams and how pay-per-interaction mechanics can be operationalized as the monetization channel for the deception.
Enterprise Customer Service And Interactive Media: OSENSE (光禾感知科技) was presented as focusing on virtual being–related enterprise interaction capabilities spanning virtual humans, intelligent customer service, AI image/video generation, and “spatial”/environment technologies, positioned as a stack for building deployable and scalable digital interactive experiences that raise interaction efficiency and data value by combining generated imagery with virtual-human interaction workflows.
E-Commerce Live Streaming And Marketing Automation: Multiple items framed “digital human” livestreaming as a high-volume commercialization pathway and a regulatory risk area, including discussion of misrepresentation, exaggeration, and impersonation via AI-generated likeness and voice; PressReader’s case study reported that the cleaning brand Lao Guanjia (老管家) operated a Kuaishou (快手) storefront where fewer than ten digital humans were credited with materially increasing daily earnings, highlighting operational scaling through multi-avatar staffing; a patent filing was reported by CEC Media (中汽传媒) and CEC Information Technology (中汽信息科技) for a marketing-oriented system that automatically generates and manages AI content and digital humans from requirement intake through delivery; and commentary tied these practices to newly strengthened live e-commerce supervision that explicitly brings AI-generated content such as digital-human anchors into scope, emphasizing accountability for synthetic presenters and the market incentives driving their adoption.
Fashion, Retail Identity, And Virtual Shows: Aokang (奥康) was reported to have introduced a virtual digital human named Aoki as part of its digital transformation and then staged an AI virtual runway/show-style event positioned as moving from single-point experimentation to more systematized use of digital humans in product and brand presentation, treating the synthetic character as a persistent brand-facing spokesperson rather than a one-off campaign asset.
Healthcare, Digital Twins, And “AI Doppelgängers”: Several entries used “AI avatar/digital double” language for health contexts, including references to large-scale deployment of clinician “AI doubles” inside consumer medical apps and remote care workflows (without naming specific doctors), the use of digital-human style products within hospital settings as part of “AI+medical” tool portfolios, and a Zhejiang case in which a traditional Chinese medicine institution reportedly built an interactive digital human around a named specialist for sleep-disorder services to support consultation-like interaction; in addition, non-China items referenced AI “digital biomarker” and sensor approaches being leveraged toward remote therapeutic monitoring use cases, which, while framed as “digital life”/“digital biology,” functioned in practice as AI-generated physiological/pose signals and monitoring outputs that can underpin companion-like guidance or clinician-facing decision support.
Education And Knowledge Delivery: Shandong Mantianxing Information Technology (山东满天星信息科技有限公司) was reported to promote an “AI digital human teacher” concept that clones a teacher’s appearance, voice, and motion into a cloud-based synthetic instructor, coupled with a self-updating subject knowledge graph, positioning the digital human as a scalable instructional presence that can be replicated across classrooms and labs; Jiangsu provincial “AI+” policy coverage also cited educational digital-human IP characters used for explaining knowledge points in public-facing demonstrations, framing avatars as standardized communicators for pedagogy and civic learning.
Public Services, Culture, And City-Level Deployment: Regional and municipal items described public-sector and cultural deployments including a Nanjing education agency’s digital-human IP characters used as interactive explainers, Guangxi cultural-and-tourism public service showcases featuring holographic digital humans as experiential exhibits, and a Guiyang museum renewal that explicitly referenced a “Wang Yangming digital human” as part of exhibit modernization; city governance and international-visitor service narratives also referenced an AI digital human named Alice used to deliver multilingual support services, illustrating synthetic characters as front-desk equivalents in volunteer coordination and urban service delivery contexts.
Entertainment Creation, Short-Form Content, And “Digital Double” Experiences: JD.com (京东) was reported to launch an AI film/short-video creation contest that explicitly incorporated a JD digital human and an IP character as creative elements while allowing broad freedom in tool choice, and market/industry items described “digital doubles” as applicable across live commerce, virtual teachers, film/TV performance, and metaverse experiences; separately, corporate collaboration news indicated Huanrui Century (欢瑞世纪) working with StepFun (阶跃星辰) to build a joint AI lab and develop agent-based short-drama products, aligning synthetic-character pipelines with automated narrative production and interactive entertainment formats, while Gigabyte (技嘉) was described as using a guided “digital double” experience to lead participants through interactive AI product journeys and generate a personalized recap video as a commemorative output.
Corporate Strategy And Virtual-Human Industrialization In China: Tianyu Digital Technology (天娱数科) was framed in investor-oriented coverage as emphasizing virtual-human commercialization, including claims of substantially reduced production costs through batch production and a high-fanbase flagship “Guofeng” virtual human named Tianyu (天妤), positioning virtual beings as a revenue-driving product line rather than an experimental showcase; additional local-industry reporting highlighted Chengdu Digital Sky Technology (成都数字天空科技有限公司) building an ultra-realistic digital-human modeling and animation performance pilot platform, reflecting continued investment in production infrastructure for high-fidelity avatars and synthetic character performance.
January 14 News
Education and Public Services: Chongqing Xinhua Publishing Group (重庆新华出版集团) is linked to an AI education strategic cooperation with Fengdu County in Chongqing, positioned alongside the view articulated by Baidu (百度) executive Wu Chenxia that digital humans are a key carrier for human–computer interaction in an AI-native era, while a concrete classroom case in Xi’an shows an AI digital human issuing an invitation for students to become “Chinese Wisdom Youth Story Narrators,” using the digital human as a pedagogical prompt to drive participation and address teaching friction points through narrative-style engagement rather than conventional instruction.
Healthcare: Sichuan Tianfu Bank (四川天府银行) applies a multimodal foundation-model approach in its “Tianfu AI Micro-SME Loan” workflow, explicitly using a digital-human interview step, intelligent due diligence, and automated approval to compress micro-loan processing from days to much shorter cycles; in Hangzhou, an occupational-health AI digital human named “Zhifang Yunpeng” functions as a public-health explainer intended to improve awareness in a more vivid, accessible format; in Guangxi, a digital-human medical agent called “Miyu Yitan” is described as an AI consultation system that iterated into a multilingual ASEAN-capable version, emphasizing fast and accurate Q&A-style triage or guidance; and Meditech (麦迪科技) describes consumer-facing medical and wellness products that include medical report and document interpretation, online AI-assisted diagnosis support, “AI doctor clones,” wearable-integrated health management, hospital–out-of-hospital data connectivity for full-course health management, and tailored services for specific populations, while the health platform Antifu (健康AI蚂蚁阿福平台) is associated with a reported pattern of large-hospital doctors opening their own “AI clones” to provide 24/7 consultation coverage at scale.
Entertainment and Virtual Production: AOTO Electronics (奥拓电子) is tied to the activation of a 3,000㎡ virtual production studio in Yongchuan, Chongqing, described as integrating a core display system plus real-time digital human technology, AI image segmentation, and ST2110-based workflows, indicating a pipeline aimed at live compositing and broadcast-grade IP video transport; separately, Shiyou Technology (世优科技) is profiled as a “full-stack” AI digital human vendor with highlighted deployments including a prominent staged digital-human performance linked to the Hangzhou Asian Games, used as an emblematic case of high-visibility, large-event digital-human presentation.
Marketing and Commerce: China Telecom (中国电信) is associated with an AI digital-human e-commerce livestream training base in Shufu County, where practitioners use digital humans to sell local “Jiangguoguo” products and to operationalize livestream commerce skills; Keyiyun (客易云) is framed around “digital human AI livestream” as a route to higher-efficiency, more automated live marketing; Kuaishou (快手) is the platform context for the cleaning brand Laoguanjia (老管家), which is described as expanding to nearly ten stores and using fewer than ten digital humans to materially boost daily earnings, presenting digital-human livestreaming as a low-marginal-cost sales amplifier; and the regulatory environment is described as tightening to require clearer identification of digital-human anchors as AI-generated content, explicitly motivated by risks including merchants generating authoritative-looking “elder expert” avatars and then funneling older users into private channels to induce high-priced health-product purchases.
Enterprise and Industrial: Jiangsu Cable (江苏有线) describes an “AI portal” upgrade that includes a “digital human platform” offering two operational modes—scripted/speaking (for prepared delivery) and interactive (for dialogue)—positioning digital humans as a front-end layer for internal productivity and office workflows; Huawei (华为) and Yunnan Construction Investment Logistics (云南建投物流) jointly present a “smart logistics” benchmark and launch R&D on a logistics foundation model named “Yunlian Logistics Large Model,” situating digital-human and agentic capabilities within supply-chain digitization discussions; in industrial policy framing, the Ministry of Industry and Information Technology’s direction is summarized as encouraging platforms to cultivate high-value scenarios such as process-automation assistants and smart-inspection digital humans as part of industrial internet and AI fusion; Diweixun (ST迪威迅) is linked to a “rapid digital human generation system” described as enabling efficient creation of diverse virtual characters to support digitized interactive experiences; and in the set of officially recognized metaverse “typical cases” in Nanjing, Nanjing Lopu (南京洛普股份有限公司) is associated with an LED display system for immersive metaverse cultural-tourism scenarios, Silicon Intelligence (南京硅基智能科技集团股份有限公司) with an AIGC digital-human intelligent interaction platform positioned as a cross-industry metaverse interaction tool, and Depth Vision (纵深视觉科技（南京）有限责任公司) with light-field naked-eye 3D for immersive remote operation in mining, collectively indicating that digital humans are being packaged with display, interaction, and 3D visualization stacks for deployable sector solutions.
Cultural, Tourism, and Heritage: A tourism digitization case references an online smart-tourism platform called “Xia Yangzhou” that uses an AI digital human for bidirectional voice–text conversion to support user interaction and accessibility; in Shanghai, a digital guide framed as a “bowl of human warmth” is distributed through a WeChat entry point to provide a one-stop intangible-cultural-heritage New Year experience, treating digital content and guided exploration as a cultural-transmission mechanism; in Guizhou, a “Wang Yangming digital human” is presented as part of a memorial-site renewal to enhance interpretation of Wang Yangming-related content; in vocational training, a Guizhou nursing college describes developing multiple VR emergency-tech software packages that integrate an AI intelligent teaching assistant and a digital human component oriented to practical first-aid scenarios; and in retail-experience infrastructure, Tongxiang Intelligent (同享智能（北京）科技有限公司) is cited as an example of supply-chain support for immersive retail storytelling that includes interactive virtual personas and triggered animated experiences, while Zhanghui Technology (掌慧科技) presents a mascot that is framed as a “virtual employee” designed to function inside the service chain rather than remain a superficial brand endorser.
January 13 News
Healthcare: Health AI Ant Afu (蚂蚁阿福) was reported as a platform where more than 500 doctors from top-tier “Class III Grade A” hospitals have opened personal “AI doubles” that can stay online around the clock to extend physicians’ service reach, framed alongside survey findings that many doctors already use AI tools and that most respondents are optimistic about AI-assisted medicine; separately, Douyin Xinghui (抖音星绘) was described as an app feature for creating an “AI double” via an upload-based workflow, presented as a consumer-facing creation tool rather than a clinical system.
Education: In Xi’an, a classroom example described an AI system turning the classical Chinese text “Sima Guang” into a song and then using a digital human to issue “Smart Youth” tasks that guide fourth-grade students through deeper exploration of the text’s meaning, positioning the digital human as an instructional guide and activity dispatcher embedded in everyday teaching.
Cultural Tourism: A recurring case was Yangzhou’s “Xia Yangzhou” online smart tourism platform (“下扬州”线上智慧旅游平台), described as using an AI digital human with voice–text mutual conversion to deliver accessible guidance such as tours and maps for older travelers with hearing difficulties and to provide personalized recommendations based on individual interests; in Jiangsu, the “Zaolinwan Cultural Tourism” (枣林湾文旅) new-media publicity matrix launch highlighted a digital human character “Zaolinger” (枣灵儿) as an interactive, attention-driving presence for regional tourism communication; in Guangdong, an “AI digital human Li Bai” interactive themed show was described as a New Year holiday collaboration between the NuwaAI team (NuwaAI) and Gankeng Ancient Town (甘坑古镇), emphasizing live, on-site interaction as the core attraction.
Government And Public Services: In Sichuan’s Linshui County, a “Digital Human Social Security” (数字人社) 24-hour self-service area was reported as opening to the public, positioning the digital-human interface as a front-end for always-available government services; at transport hubs, a Guangxi airport case described Shiyou Technology (世优科技) deploying its “Bota” AI digital human product (波塔) at information counters, self-service kiosks, or other terminals to provide 24/7 passenger assistance including flight queries, route guidance, facility introductions, and multilingual translation; in Fujian, the Fuzhou Women’s Federation (福州市妇联) was reported as building an AI digital human team for “smart digital advocacy,” anchored by child-friendly IP characters “Rongrong” and “Momao” (榕榕茉茉) alongside role-based personas such as hosts and lawyers to deliver policy and civic messaging in a more scalable, repeatable format.
Enterprise And Industry Tools: Nanjing Silicon Intelligence Technology Group (Nanjing) Co., Ltd. (南京硅基智能科技集团股份有限公司) was cited in connection with an “AIGC digital human intelligent interactive platform” (AIGC数字人智能交互平台) framed as enabling a new metaverse interaction paradigm across multiple industries; Depth Vision Technology (Nanjing) Co., Ltd. (纵深视觉科技（南京）有限责任公司) was cited for an immersive remote-operation scenario in the mining industry, presented as part of the same “typical metaverse cases” context; in Shandong, Qingdao Qianrui Shulian Information Technology Co., Ltd. (青岛千瑞数联信息技术有限公司) was cited for a customer-service digital human named “Xiaorui” (客服数字人“小瑞”), presented as a representative “typical digital human” case; Shandong Jindong Digital Creative Co., Ltd. (山东金东数字创意股份有限公司) was reported as being selected in a “typical metaverse cases” product/application category, indicating recognized industry application but without additional functional detail in the provided excerpts; Jiecheng Co., Ltd. (捷成股份) was reported as developing B2B AI applications for broadcasters, film/TV, short-video producers, and MCN operators, explicitly including an “AI virtual host” capability (AI虚拟主持人) alongside AI text/script generation and AI flat-asset generation as part of its toolchain.
Marketing And Live Commerce: Digital human anchors were described as entering rural e-commerce as “never-tired new farmers,” with one reported case claiming a digital human streamer achieved RMB 190 million in annual sales while being positioned as a low-cost, high-output solution for product marketing and livestream operations, and the broader framing emphasized that digital humans can substitute for scarce labor and maintain continuous selling cycles; a separate market-oriented narrative noted investor attention to the “virtual digital human” concept sector, with references to publicly traded firms such as Fushi Holdings (福石控股) and Chuanwang Media (川网传媒) appearing as sector constituents in market summaries rather than as described builders of a specific digital-human deployment in the excerpts provided.
Entertainment And Media: Zhejiang Television (浙江卫视) was described as presenting a stage lineup that included multiple “digital human” appearances—its own “Gu Xiaoyu” (谷小雨), Jay Chou’s “Zhou Tongxue” (周同学), and the “White Lady” (白娘子) from “White Snake” (《白蛇》)—explicitly framed as AI-driven stage manifestations that add interactive, high-tech spectacle, with the same entertainment coverage also referencing large-scale 3D interaction elements tied to game IP without detailing additional synthetic-character capabilities.
Regulation And Governance: Reporting on livestream e-commerce governance stated that “digital human anchors” and other AI-generated content forms have been brought within the scope of platform or regulatory oversight, positioning synthetic presenters as regulated subjects of content management rather than as novelty performers, and emphasizing that supervision frameworks are being updated to account for AI-generated hosts operating at scale.
January 12 News
Healthcare: Kaiping Medical Security Bureau (开平市医疗保障局) and the Jiangmen Municipal People’s Government (江门市人民政府) described a government-service reform case in which the “Kaiping Xiaozheng” AI digital human guidance system (“开平小政AI数字人”导办系统) was put into formal use for medical-insurance (医保) services; the system reportedly ingests 35 business guides, 15 high-frequency handling notes, and common hot-issue Q&A, and is used to answer residents’ inquiries through an interactive, consultative digital-human interface oriented to front-desk guidance and self-service consultation.
Enterprise And Government Services: Shiyou Technology (世优科技) was reported as having its Bota 4.0 AI digital human “Xiya” (波塔4.0 AI数字人“希娅”) selected into the publicized list of 2025 “typical metaverse cases” jointly issued by the Ministry of Industry and Information Technology and other central agencies, positioning “Xiya” as a representative AI digital-human application within a national-level metaverse case framework; separately, a Beijing government-service scenario was described in which an AI digital-human guide named “Feng Xiaoshu” (AI数字人引导员“丰小数”) provides continuous on-site guidance through voice/text interaction in a public service hall with an always-on service posture; in port/industrial administration, Tianjin Port Information Technology Development Co., Ltd. (天津港信息技术发展有限公司) issued a tender for a “digital-human assistant” large-model application development project that explicitly includes building an agent-development platform with a general-purpose agent capability layer, indicating an enterprise-grade path where “digital human” is treated as an application-layer interface over an intelligent-agent platform rather than only as a media asset.
Marketing And E-commerce: iFLYTEK (科大讯飞) was cited for linking its Spark cognitive large model (星火认知大模型) with commercialization of its virtual being “Su Xiaomei” (虚拟人“苏小妹”), emphasizing cross-scenario IP linkage to extend brand co-branding and culture-and-tourism (文旅) collaborations as a “technology + content + scenario” marketing ecosystem; Tianyu Digital Technology (天娱数科) was described as operating the virtual being “Tianyu” (虚拟人“天妤”) with a reported fanbase exceeding 30 million and using endorsement-style deployments in beauty and jewelry while monetizing through a mix that includes live-stream revenue sharing and licensing/authorization structures tied to its AI marketing services; Baidu (百度) was reported to have launched in Macau an e-commerce solution centered on Huibo Xing digital-human livestreaming (慧播星数字人直播) as the core of an end-to-end merchant growth package aimed at customer acquisition and small-business scaling, framing the digital human specifically as the live-commerce front-end for conversion rather than a standalone character product.
Entertainment And Content Creation: ByteDance (字节跳动) was reported to be pushing an AI video application that foregrounds “AI avatar/digital double” creation (AI分身/数字分身) via image-based generation and “intelligent co-shooting” style workflows, with coverage noting that its Doubao app (豆包) supports template-style AI video play but does not yet support users “personally performing” via an AI double, instead limiting creation to photo-based AI portrait outputs; in Taiwanese charity promotion coverage, entertainer Wang Tong (王瞳) was described as appearing alongside her AI double (AI分身) and leading the AI girl group “DailyNeo” (AI女团“DailyNeo”) in a public-welfare campaign format, using the synthetic persona presentation to deliver scripted promotional speech for donated meal fundraising, illustrating a use case where an AI double functions as a performative media surrogate in public communications.
Gaming: Lilien Games (莉莉恩工作室) was reported to be releasing a VTuber-themed courtroom deduction game “Who Is The Operator?” (《谁是中之人》) on Steam, explicitly building gameplay around the “operator behind the VTuber” (中之人) concept and pairing it with demo-based suspect-prediction/character-popularity discussion, which positions “virtual being” culture as the narrative premise rather than using a digital human as a service interface; in mobile game collaboration coverage, the game “Blind Box Party” (《盲盒派对》) was reported to have launched a limited-time crossover with the virtual UP host Anqing Ankii (虚拟UP主安晴Ankii), adding themed content built around a “penguin girl” character skin/reward structure and demonstrating a common case where a virtual creator identity is integrated into game content as an event IP asset.
Standards, Governance, And Risk Controls: Securities Times (证券时报) reported the release of China’s first national standard for virtual digital humans, titled “General Technical Requirements for Virtual Digital Humans” (《虚拟数字人通用技术要求》), framing it as filling a gap in technical norms and providing unified requirements and evaluation criteria specifically for customer-service-type virtual digital humans, which formalizes baseline expectations for R&D, production, and deployment; chengdu.cn commentary on livestream governance described AI cloning and digital-human livestreaming as emerging marketing forms that also create deepfake and personality-rights infringement risks, and stated that new regulatory moves set two “red lines” focused on authenticity and authorization, explicitly prohibiting the use of AI to forge another person’s likeness without authorization in livestream commerce contexts.
Infrastructure And Tooling For Virtual Being Production: Multiple reports tied virtual-being generation and operation to compute and content pipelines rather than character design alone, describing film VFX rendering, 3D virtual-human generation, and large-scale online game operation as typical cultural-and-creative workloads that depend on high-performance GPU compute, alongside AI training and scientific simulation workloads that compete for the same compute substrate; in a regional “high-tech product promotion” activity in Chongqing, a “digital human full-process solution” (数字人全流程解决方案) was described as providing technical support for metaverse and VR content production, indicating an integrated production workflow approach where digital humans are treated as pipeline outputs spanning asset generation, animation/presentation, and scenario deployment.
January 11 News
Ethical Contexts: Shanghai’s advertising rules treat “digital virtual humans” used in ads as a regulated advertising technique, requiring explicit authorization when a digital virtual human uses a real person’s likeness, with the authorization specifying scope such as ad type, distribution platform, and usage period, and prohibiting unauthorized use of another person’s portrait or voice, framing digital humans as a compliance and rights-management problem that combines identity rights, synthetic media production, and platform distribution controls.
Marketing: BlueFocus (蓝色光标) positions virtual humans alongside virtual objects, virtual spaces, and XR studio production as part of an end-to-end marketing and marketing-technology service chain delivered across major global markets, while emphasizing commercialization of virtual-human capabilities as a revenue-generating deployment rather than a demo feature, and it deepens its collaboration with Volcano Engine (火山引擎) around these virtualized marketing assets and production workflows.
Healthcare: Ant Group (蚂蚁集团) presents “Afu” as a health-service system built on a proprietary domain foundation model and co-created with academicians and clinical experts, connecting large-scale hospital resources and real doctors while also organizing “AI doubles” of named physicians for professional Q&A and extended service coverage, and Shanghai’s western pediatric consortium operationalizes a child-facing cartoon digital human persona (“Tong Baobao”) as the front-end for pediatric health management functions spanning health education, structured check-up experiences, and service navigation within a multi-institution care network.
Enterprise and Public Services: Shiyou Technology (世优科技) concentrates on a full-stack AI digital-human pipeline that supports rapid avatar customization and interactive deployment, demonstrated through a government-service hall guide digital human (“Feng Xiaoshu”) providing continuous 7×24 voice-and-text guidance for in-person administrative services, through airport-style scenarios where a digital human provides multilingual flight queries, wayfinding, and facility introductions as an always-on service layer, and through enterprise-showroom implementations where a brand-mascot digital human (“Jibao”) also acts as an intelligent explainer in logistics exhibition spaces and where corporate exhibition halls use “digital human employees” for guided tours; SenseTime (商汤) is positioned as an enabling platform layer for city governance via a planning-and-resources AI model supported by its infrastructure offering (“SenseTime Large Device,” 商汤大装置), aligning digital-human interaction with broader municipal AI systems rather than treating it as a standalone avatar feature; JD.com (京东集团) elevates digital humans into a dedicated business unit (“Chameleon,” 变色龙) tasked with productizing multi-modal AI and digital-human capabilities across JoyAI App and embedded solutions such as JoyInside, and it extends experimentation into brand collaborations that explicitly combine multi-modal AI with digital-human interaction in consumer-facing deployments.
Entertainment and Media Production: Ningmeng Shuguo (Chongqing) Technology (柠萌数果（重庆）科技) promotes a “real-time motion-capture 3D digital human full-process solution” aimed at metaverse and virtual-reality content production, centering on performance capture, character driving, and end-to-end pipeline efficiency for producing controllable 3D digital performers; Tianjin Daily (天津日报) describes a broadcast program format that puts real performers and digital humans on stage together using AIGC and virtual production (VP) to synchronize performance, compositing, and scene generation, treating the digital human as an on-stage co-performer rather than a post-production visual effect.
Cultural and Tourism Contexts: NuwaAI (NuwaAI团队) collaborates with Gankeng Hakka Town (甘坑古镇) to deploy an AI digital human embodiment of “Li Bai” as an on-site cultural IP experience, framed as a location-based digital character that merges heritage tourism presentation with interactive AI persona design, with technical support attributed to Guangdong Southern Smart Media Technology (广东南方智媒科技有限公司), indicating a production-and-operations stack that ties a recognizable synthetic character to a specific destination venue.
Education and Skills: Major arts institutions including Communication University of China (中国传媒大学), Beijing Film Academy (北京电影学院), Shanghai Theatre Academy (上海戏剧学院), and Nanjing University of the Arts (南京艺术学院) expand admissions-oriented training pathways aligned to metaverse, virtual humans, and immersive experience industries, treating virtual-human production as a professionalized creative-technical discipline; Beijing Zhongjiao Zhixun Equipment Co., Ltd. (北京中教智讯设备有限公司) packages vocational training environments that explicitly include a “digital human e-commerce system” module and associated sentiment-analysis practice, integrating locally deployable foundation-model options such as DeepSeek (深度求索) and Wenxin Yiyan (文心一言) into applied curricula designed to support digital-human-driven commerce operations and content workflows.
January 10 News
Healthcare: In Chongqing, an unnamed “…(Chongqing) Technology Co., Ltd.” released a brain–computer interface hemiplegia rehabilitation solution that combines biosignals with rehabilitation engineering, positioned as a practical clinical-rehab application rather than a generic demo, while Ningmeng Shuguo (Chongqing) Technology Co., Ltd. (柠萌数果（重庆）科技有限公司) released a real-time motion-capture 3D digital human capability positioned for full-process, live, real-time performance. In Beijing, ThunderSoft (中科创达) deployed a “smart guidance” 3D simulated female guide digital human at Peking University First Hospital, described as real-time rendered with natural motion and used to shift navigation from “people searching for services” to “services following people.” In hospital-facing consumer guidance, Ant Group (蚂蚁集团) was cited in relation to “Ant A’Fu” (螞蟻阿福), where named doctors’ “AI avatars” were described as answering large volumes of health consultations, framing the virtualized clinician persona as a scalable front-end for triage and routine Q&A rather than a replacement for in-person care.
Enterprise: In Guangxi, Umetrip (航旅纵横), developed under TravelSky (中国航信), launched an intelligent customer-service digital human integrated into an all-in-one kiosk (“Guifeifei,” 桂飞飞) that entered live operation at Nanning Wuxu International Airport, presenting the digital human as a frontline service interface for airport passengers. In Jiangsu (Suqian, Suyu District), “AI digital human + government services” was described as a 7×24 “chat-and-handle” model intended to let users converse while completing administrative tasks, positioning the digital human as an always-on guided transaction layer for public services. In Tianjin, Tianjin Bank (天津银行股份有限公司) issued procurement for upgrading its intelligent customer-service platform to add a digital human and “intelligent Siri function,” indicating a bank call-center and online service workflow that uses a humanlike interface to reduce waiting, route queries, and standardize responses.
Entertainment: Zhipu AI (北京智谱华章科技有限公司) was reported in connection with a Hong Kong listing narrative that explicitly framed digital humans as connectors between virtual and real contexts and as “core carriers” for intelligent communication, emphasizing a shift from pure entertainment “traffic” use toward broader roles; separately, JD.com (京东) was reported to have created a “Chameleon” business unit responsible for building and commercializing core AI products including the JoyAI app, JoyInside, and digital humans, signaling an internal product-line consolidation where synthetic characters and digital-human interfaces are treated as monetizable platform capabilities rather than one-off marketing assets.
Marketing: Shanghai Ren Shijian Digital Technology Co., Ltd. (上海人食间数字科技有限公司) was described as using AI virtual digital humans to generate “old expert” and “host” personas for short-video advertising targeted at middle-aged and elderly audiences, coupled with therapeutic claims about health products, presenting a concrete case of synthetic character deployment for precision ad delivery and persuasive authority simulation.
Gaming: A China-outbound industry narrative described the pairing of AI with digital human technology as a response to diminishing returns from traditional localization and marketing in game development and publishing, presenting digital humans as a practical tool for overseas growth loops through more interactive promotion, localization-adjacent presentation, and user engagement amplification, without naming a specific vendor in the provided excerpt.
Cultural: In Anhui (Hefei, Luyang District), culture-and-tourism development was described as incorporating frontier elements including digital humans into night-time scenic experiences and performances, indicating a direction where cultural venues integrate virtualized presenters or performers into public-facing entertainment and documentation-driven visitor experiences. In education contexts tied to local culture, Changsha classroom practice was described as using an AI digital human as a “tour guide” to support ideological and civic education through child-friendly narrative delivery, while a Hainan classroom example described designing a history-lesson digital human (“Zhang…”) to make traditionally dry content more engaging, positioning the digital human as a role-based explainer embedded inside lesson flow.
Ethical Contexts: National rulemaking and enforcement narratives repeatedly treated “digital human livestreaming” as a distinct governance object, with the State Administration for Market Regulation (国家市场监督管理总局) and the Cyberspace Administration of China (国家互联网信息办公室) described as bringing AI-generated content, including digital human streamers, into supervision frameworks that add traffic-control tools, require identification or labeling of AI-generated content in livestream rooms, and tighten platform responsibilities around rule-setting and merchant rights, explicitly linking digital humans to risks such as deepfakes, AI impersonation, and misleading marketing; parallel market commentary in Hong Kong financial coverage referenced the regulatory impact on platform firms including Alibaba (阿里巴巴) and JD.com (京东), framing digital human streamers as a compliance category that platforms must operationalize through governance processes rather than treating as purely creative or growth features.
January 09 News
Healthcare: Changzhou First People’s Hospital (常州市第一人民医院) is described as deploying an “intelligent digital human” (智能“数字人”) to deliver clearer hospital navigation and guidance, positioning the virtual being as a front-line service interface that handles wayfinding and patient routing functions in a clinical setting.
Marketing And Commerce: JD.com (京东) is described as accelerating commercialization of its virtual-being stack by centralizing core AI products including JoyAI App (JoyAI App) and JoyInside (JoyInside) alongside “digital humans,” indicating a productized pathway that combines language, speech, vision, and digital-human presentation for consumer-facing interactions and embedded experiences; Shanghai Ren Shijian Digital Technology Co., Ltd. (上海人食间数字科技有限公司) is described as using AI virtual digital humans (AI虚拟数字人) to generate “senior expert” and “host” personas for short-video advertising and targeting older audiences, with the key operational detail being persona fabrication for persuasive health-product marketing; Jisi Technology (集思科技) is described as operating a large-model digital-human livestream platform named “Xiaoxiao” (“销销”大模型数字人直播平台) intended to address practical constraints in e-commerce livestreaming; sector signals are also reflected in market coverage that references the “virtual digital human” segment and highlights Yiwang Yichuang (壹网壹创) and Hand Enterprise Solutions (汉得信息) as leading movers within that theme, framing them as public-market exemplars tied to the broader commercialization narrative around virtual digital humans.
Cultural And Tourism: A Shandong exhibit context describes a photoreal 3D “Confucius” digital human (孔子3D超写实数字人) achieving a 4K, film-grade effect (4K影视级效果) and being presented alongside the “Haoyu” cultural-tourism digitalization platform (“浩宇”文旅数字化平台), treating the virtual being as a high-fidelity cultural figure for tourism-facing presentation and place-based digital experiences; additional regional proposals and civic demonstrations describe digital humans as hosts or guides for immersive cultural-tourism scenarios and ceremonial participation, including scripted leading of group activities in public-institution settings as part of digitally mediated cultural engagement.
Gaming And Sports Entertainment: Lenovo (联想) is described as building high-precision avatars (高精度Avatar/“虚拟人”) for well-known football players to more accurately reproduce match details and support decisions around key calls such as offside, positioning the avatar as a realism and analytics layer tied to sports presentation; Tencent Games (腾讯游戏) is described as being involved in a “digital empowerment” approach to intangible cultural heritage transmission, where an “original digital human” (原创数字人) is used as an interactive narrative or guidance layer within a heritage context, linking avatar-mediated storytelling to cultural continuity.
Enabling Technologies And Production Pipelines: The National Supercomputing Center in Jinan (国家超级计算济南中心) is presented as a core computing node within a provincial “computing power network” narrative, supplying the infrastructure backdrop for high-fidelity digital-human production and related large-scale digital applications; the University of Macau (澳门大学) is referenced in connection with a breakthrough in real-time facial animation that reduces stutter and latency for virtual anchors, framing the key technical bottleneck as real-time facial performance driving for synthetic characters used in livestreaming and broadcast-like scenarios; Inner Mongolia YuanZhi PuDao Technology Co., Ltd. (内蒙古源智普道科技有限公司) is described as providing an “AI platform” that can convert an entire textbook into a complete instructional package including PPT, digital-human videos, and a question bank, using the digital human specifically as an automated presenter within a scalable content-generation pipeline.
Governance, Compliance, And Risk: Haihe Media Center (天津海河传媒中心) is referenced in connection with reporting that digital-human livestream selling should have its “identity” explicitly labeled, emphasizing disclosure of synthetic-host status in commerce-facing broadcasts; multiple national-level policy summaries describe regulatory moves that explicitly bring “digital human anchors” (数字人主播) and other AI-generated content into livestream e-commerce governance, requiring identification/labeling to prevent virtualized impersonation and to reduce the use of synthetic characters as a shield for false or misleading marketing; risk reporting also describes “AI face-swap + virtual persona” (AI换脸+虚拟人设) fraud as an emerging scam pattern, where synthetic identity construction functions as the social-engineering layer that enables financial loss, reinforcing the practical compliance need for provenance signals, platform enforcement, and consumer-facing disclosure around AI-generated personas.
January 08 News
Governance And Regulation: Several reports describe a regulatory move to bring AI-generated livestream commerce content explicitly under oversight, with “digital human” livestream anchors (数字人主播) treated as a regulated form of AI-generated content in livestream e-commerce. In a January 7 report carried by multiple outlets, the head of the State Administration for Market Regulation’s online-market supervision function, Zhu Jianqiao (朱剑桥), said the State Administration for Market Regulation (市场监管总局) and the Cyberspace Administration of China (国家网信办) had recently jointly issued rules for livestream e-commerce that add digital-human anchors and other AI-generated content into the regulatory scope, alongside tools such as traffic/flow governance requirements and long-term retention expectations for livestream marketing records (including multi-year preservation obligations referenced in Shanghai-related coverage). Separate Shaanxi coverage links the same rules to consumer protection goals such as preventing platforms from using “big data price discrimination” (“杀熟”), explicitly noting that digital-human anchors and other AI-generated content are included so that new technical forms do not become a channel for misleading, unfair, or evasive practices in livestream selling.
Marketing And Livestream Commerce: JD.com (京东) is repeatedly cited for a specific “digital human livestream-room replica” capability, described as a way to convert an existing real-person livestream into a high-fidelity digital counterpart: merchants upload recorded livestream video assets and quickly generate a “digital avatar/digital double” (数字分身) that reproduces the host’s appearance, voice, expressions, and the livestream-room set with all-dimension, high-fidelity reconstruction, and the feature is described as available to JD platform merchants through the JD Merchants system and related service channels, including references to applying via the JD platform and subscribing through the Jingmai service marketplace (京麦服务市场). County-level and regional commerce reporting also frames “AI digital human livestreaming” as a way to maintain 24-hour continuous sales promotion for local products by combining e-commerce channels with always-on digital presenters, positioning digital humans as a practical method for extending livestream capacity beyond human scheduling limits while keeping a recognizable host identity.
Enterprise And Industrial Ecosystems: Regional industry reports describe enterprise clusters and supplier ecosystems forming around digital-human R&D and deployment. In Yinchuan’s industrial park context, Ningxia Lingji Technology Co., Ltd. (宁夏灵机科技有限公司) and Yingwei Intelligence (英炜智慧) are named among companies gathered into an “AI ecosystem community,” with coverage indicating the local enterprise set spans digital-human R&D as part of a broader computing-power and AI industrial base. In Zhejiang-focused coverage, Yunjiazu (Zhejiang) Technology Co., Ltd. (云家族（浙江）科技有限公司) is explicitly associated with AI, digital-human development, and “metaverse” related work, and was reported as having obtained national high-tech enterprise recognition, implying formal validation of its R&D profile. Capital-markets and corporate-development reporting also flags Super Fusion Digital Technologies Co., Ltd. (超聚变数字技术股份有限公司) entering a public listing辅导 process with CITIC Securities (中信证券) as the辅导机构, with the surrounding context treating “virtual digital human” as a notable thematic category within broader tech-sector positioning.
Media, IP, And Virtual Hosts: Entertainment and media-facing use of synthetic characters is illustrated by the appearance of an AI idol as a stage host in a major knowledge-platform ceremony: Baidu (百度) is referenced via coverage of a Baidu Baike annual ceremony where the AI idol “Su Xiao” (苏晓) served as a “cross-dimensional” virtual host, emphasizing a format where a synthetic character is presented as the on-stage presenter rather than a human MC. Separately, a patent-focused report says Hangzhou Yuanmei Technology Co., Ltd. (杭州元媒科技有限公司) filed an application for a “virtual digital human” media-display method and system for converged media, positioned as a technical approach intended to provide a user experience different from conventional information-promotion methods, aligning virtual digital humans with structured media presentation and distribution systems rather than single-platform performances.
Education And Public-Sector Service: Multiple items describe digital humans used as service-facing “front ends” for education, guidance, and civic interaction. Nanjing’s preschool education initiative introduces an “intelligent agent” called Ning Xiaozhi (宁小智), led by the Nanjing Educational Technology Center (南京市电化教育馆), built around two digital-human IP characters, “Rongrong” (荣荣) and “Pi Xiaoshou” (辟小兽), and described as leveraging nearly a decade of curated education resources while aiming for lightweight interaction and broad platform compatibility. Higher-education and conference service coverage describes the “intelligent digital human” Long Xiaoqiao (龙小侨) acting as an AI concierge/guide at a service desk to welcome guests and support navigation and interaction in an event setting, presenting the digital human as a staffed touchpoint that can converse and assist attendees in real time. University-student support coverage in Hunan describes “AI counselor” deployments in which a digital-human counselor persona, “AI Jie Ge” (AI捷哥), modeled on a real counselor, answers student questions (for example, suggesting directions and example links for coursework or ideological-practice report topics) within seconds, framing the digital human as an always-available advisory interface rather than a generic chatbot.
Healthcare And Mental Health: One of the clearest case-study-style deployments appears in Yunnan, where Yunnan University is reported to be advancing the digital transformation of mental-health services by building multi-modal AI interaction into diversified virtual environments that include digital humans, with the stated functional role of helping users complete intervention workflows for emotion management and cognitive restructuring. The same coverage states that the resulting product has already been applied within a province-wide digitalization project for a mental-disorder census/survey (云南省精神疾病普查数字化项目), indicating an operational public-health use case that connects digital-human guided interaction with standardized screening, intervention support, and large-scale program delivery rather than one-off demonstrations.
Cultural And Tourism Contexts: Tourism and cultural-heritage coverage presents digital humans as interpretive guides and conversational reconstructions of historical figures tied to specific sites and exhibitions. A Shanxi report on a cultural exhibition says visitors can converse inside the hall with an ultra-realistic virtual digital human representing the site’s founder, Dongming Chan Master (东明禅师), implying an experience design where a named historical persona is rendered as a conversational synthetic character for interpretation and engagement. Wuhan local policy-proposal coverage similarly frames digital humans as a way to create immersive tourism scenes, using the concrete example of encountering a poetry-reciting “Li Bai” (李白) at Yellow Crane Tower (黄鹤楼) who can explain history, positioning the virtual persona as an on-demand, place-based cultural interpreter designed to differentiate city tourism experiences through interactive storytelling.
January 07 News
Finance And Markets: Reports on mainland China’s A-share trading and sector rotation describe “virtual humans” and “virtual digital humans” as part of the concept/theme baskets that strengthened alongside areas like brain–computer interfaces, insurance, storage, CRO, analog chips, and genetic testing, including coverage of the Shanghai Composite’s rare twelve-session winning streak and broad market breadth with many limit-up stocks; these items treat virtual beings primarily as an investable “concept” category rather than detailing specific implementations or products.
Enterprise And Platforms: A report on 2025 large-model project bid results highlights “intelligent customer service and virtual humans” as standout winning scenarios, and characterizes Baidu (百度) as a comparatively rare domestic vendor with full-stack AI positioning tied to Kunlunxin (昆仑芯), Baidu AI Cloud (百度智能云), the ERNIE large model (文心大模型), and the Qianfan platform (千帆平台), framing virtual humans as a deployment form for enterprise-facing model applications rather than as entertainment IP.
Education: Nanjing’s city-level preschool-education agent “Ning Xiaozhi” (宁小智) is reported as an innovation project led by the Nanjing Audio-Visual Education Center (南京市电化教育馆), built around dual digital-human IP “Rongrong” (荣荣) and “Bi Xiaoshou” (辟小兽) and supported by nearly a decade of local high-quality education resources, with an emphasis on cross-platform compatibility, lightweight interaction, and more precise service delivery consistent with an “education agent + embodied digital character” approach rather than a generic chatbot.
Cultural And Public-Facing Experiences: Beijing City Library’s metaverse experience hall is reported to use an AI “Lu Xun” digital human (AI鲁迅数字人) for visitor guidance and book recommendation, explicitly presenting “Lu Xun” as a navigational and curation persona that recommends titles such as Records of the Grand Historian (《史记》), Book of Songs (《诗经》), and Dead Souls (《死魂灵》), within a broader plan to push “smart” library construction where AI-assisted services reduce friction in tasks like search and delivery.
Entertainment: A Gansu cultural story describes a digital human “Ruoshui” (弱水) as the narrative lead used to animate heritage into a contemporary format, including a large-scale romance animation “The Camel Prince and the Blackwater Girl” (《骆驼王子和黑水姑娘》), positioning the digital character as a recurring cultural IP vehicle for dramatization rather than as a one-off promotional asset.
Marketing And E-Commerce: Multiple reports focus on digital humans as performance media for cross-border and commerce workflows, including JD.com (京东) launching a “livestream studio replication” capability that is positioned as reducing high-cost modeling by enabling merchants to apply for the function on-platform and subscribe through the Jingmai service marketplace (京麦服务市场) using a free digital-human livestream service, and Keyi Cloud Group (客易云集团) promoting TikTok digital humans that pair ongoing multi-language emotional-expression model optimization with planned AR virtual-scene fusion to support immersive brand marketing in metaverse-like spaces; related coverage also frames NuwaAI (女娲AI) as a digital-human/agent platform marketed around low-cost multi-language adaptation (described as 12-language support) and cited with examples such as cross-border teams producing English and Japanese beauty-product videos and holiday-period SME adoption of digital-human livestreaming that claims measurable GMV and conversion outcomes, using digital humans as a scalable content-and-sales front end rather than as purely representational avatars.
Government, Employment, And Public Services: Reports describe digital humans being operationalized as service interfaces in public-sector and employment settings, including Quwan Technology (趣丸科技) being reported as building a large-event “digital human volunteer” system, Guangzhou Housing Provident Fund Management Center (广州住房公积金管理中心) being reported as creating an “AI digital human” alongside a new-generation cross-domain capability in its service modernization narrative, and people’s-daily–branded “Dreamlike Realm” (人民网“梦幻灵境”) in Changchun being reported as an AI performance/exhibition center where visitors follow an original digital human “Baize” (白泽) through a holographic program “Baize’s Fantastic Journey” (《白泽的奇幻旅程》), collectively treating digital beings as front-stage explainers, guides, and service agents for civic engagement.
Enterprise Workforce And Tooling: A workforce trend report references the emergence of “human–machine collaboration” roles and cites 58.com (58同城) as having developed an internal capability chain that includes AI video production and AI digital-human development/synthesis, presented as part of applied workplace tooling that can change frontline performance metrics, without detailing the specific model stack or vendor dependencies.
Risk, Misrepresentation, And Governance: A consumer-protection themed report from Chongqing discusses “senior classes” that allegedly use sales staff and even AI digital humans to impersonate “institute experts” or “sinology masters,” presenting digital humans as a vector for credential fraud and persuasive deception; separately, a marketing year-in-review item describes an industry evolution narrative in which AI virtual digital humans move from the traditional “operator behind the persona” model (“zhong zhi ren,” 中之人) toward “AI-native digital lifeforms,” while also noting an AI labeling/identification regime that is described as shifting the industry from uncontrolled expansion toward more regulated innovation, framing governance as a structural constraint on how synthetic characters are presented and commercialized.
Media, Technology, And R&D Company Signals: Stock-quote style reports reference digital-human-related concept exposure for listed firms such as Arcvideo Technologies (当虹科技) in association with multimodal AI and text-to-video themes, and 360 Security Technology (三六零) in association with “virtual digital humans,” “AI agents,” and “AI corpora,” but these items function mainly as market labeling rather than documenting specific virtual-being products, deployments, or creator-led case studies.
January 06 News
Healthcare: Changzhou First People’s Hospital (常州市第一人民医院) was described as deploying a “medical digital human” that functions as a rapid front-end triage interface by analyzing symptoms and recommending an appropriate clinical department, treating the digital human as an intake and routing mechanism intended to accelerate patient navigation through hospital services rather than as a clinical decision-maker.
Education: Beihang University (Beijing University of Aeronautics and Astronautics) was described as adopting a Hushida (互视达) digital-human all-in-one device to create an “intelligent classroom” in which a digital-human teaching assistant mediates spoken Q&A through speech recognition, delivers voice-based answers with a claimed 98% accuracy rate, and uses a transparent OLED display to present 3D teaching models, with a claimed 40% increase in classroom interaction attributable to the combined conversational and visualization workflow.
Enterprise And Commerce: Jiangsu Maike Digital Space Construction Co., Ltd. (江苏麦克数字空间营造有限公司) was identified as the producer behind a “digital human IP” used in Changzhou’s Jiangnan Global Harbor (江南环球港) commercial setting, where “digital human anchors” were positioned as the presentation and service face within an omni-domain smart marketing ecosystem platform environment, indicating an operational role for digital humans inside venue-scale commerce, promotion, and customer interaction loops.
Finance And Compliance: Agricultural Bank of China’s Jiangsu branch (农行江苏省分行) was referenced in connection with a centralized-authorization workflow that used a “digital human” in a review function within a business-process outsourcing procurement context, describing a governance-oriented use in which the digital human is treated as part of an auditable checking or review step in controlled financial operations rather than a consumer-facing avatar.
Government Services: A government-service hall in Urumqi’s economic and technological development zone (Toutunhe District) was described as deploying a “digital human” guidance clerk capable of handling inquiry-and-guidance tasks at the service counter level, framing the virtual agent as a front-desk navigator for administrative services. Separately, the Beijing Municipal Development and Reform Commission (北京市发改委) was described as using a digital-human representation of its director to present the first batch of four AI innovation districts at a major city AI meeting, positioning the digital human as a formal channel for policy communication. In Pudong (Shanghai), the smart-government plan described an integrated model combining offline intelligent counters, remote video assistance, and an online digital human, explicitly treating the digital human as one coordinated channel inside a multi-modal service-delivery architecture.
Cultural And Tourism: The Wang Meng Literature Digital Experience Hall was described as building a cultural knowledge graph from Xinjiang regional materials and using “intelligent digital humans” plus big-data visualization to translate literary narratives into a multi-sensory, interactive experience, placing the digital human in the role of narrator/guide grounded in structured cultural knowledge. The Sansu Culture Digital Experience Hall in Pingdingshan (Henan) was described as relying on AI digital humans and AR to enable immersive engagement with Sansu cultural content, again using a guided interpretive virtual presence combined with spatial overlays. In Yueyang (Hunan), a museum was described as implementing what it called the province’s first offline, physical-site AI digital-human tour-guide system to support interpretive explanation with interactive performance elements. Fujian Piaofutong Information Technology Co., Ltd. (福建票付通信息科技有限公司, 票付通) was included as a scenic-area digital transformation solution provider, while Qihu Technology (奇虎科技) was described as specializing in “AI digital human + immersive experience” with a claimed 98% interaction accuracy, placing the digital human as the engagement layer inside immersive tourism and cultural-experience products.
Public Communication And Civic Education: Hongling Education Group Yuanling Middle School (红岭教育集团园岭中学) was described as introducing an AI digital human character called “Legal Little Guardian” (法治小卫士) into a morality-and-rule-of-law class, using the synthetic character as a teaching mediator for civic and legal-literacy instruction. The Zouping People’s Procuratorate (邹平市检察院) was described as producing a “procurator digital human” to explain a three-stage fraud pattern—panic induction, isolation, and transfer—using the digital human as a didactic presenter for anti-scam education aimed at strengthening public awareness.
Platforms And Core Technology Development: Kuaishou’s Kling (快手旗下可灵) was described as upgrading toward a package that includes “Digital Human 2.0” alongside multimodal model iterations and audio-visual co-generation, positioning digital-human generation as a bundled capability inside a broader media-synthesis platform. Alibaba’s Amap (高德) was described as pursuing a “world model” and planning a product application built on that world model, presented as enabling infrastructure for environment-grounded intelligent applications rather than a named digital-human character. Yizhi Intelligence (一知智能) was described as making key breakthroughs in both voice interaction and digital humans, highlighting convergence between speech-driven interaction and embodied character interfaces. Yunxuetang Information Technology (Jiangsu) Co., Ltd. (云学堂信息科技（江苏）有限公司) was described as patenting a “digital human instruction correction method” intended to make corrected digital-human instructions fit user needs and scenario norms more quickly, emphasizing controllability and alignment of digital-human behavior to contextual constraints. Shandong Digital Human Technology Co., Ltd. (山东数字人科技股份有限公司) was referenced in securities coverage as a company situated within the digital-human sector, but without additional product-level specifics in the provided material.
January 05 News
Marketing And Commerce: Keyiyun (客易云) is presented as offering a TikTok (抖音国际版/TikTok)-oriented digital-human livestream solution for cross-border marketing, treating the digital human as a synthetic on-camera host aligned to livestream commerce workflows, while Hangzhou Lianshang Jubian Network Technology Co., Ltd. (杭州链熵巨变网络科技有限公司) is described as combining GEO optimization services with digital-human livestreaming as part of an AI application commercialization stack aimed at local exposure and conversion.
Cultural And Tourism: Shaanxi Provincial People’s Government (陕西省人民政府) materials describe Xi’an deployments where AI virtual digital humans perform Tang-poetry interactions in the “Chang’an Twelve Hours” theme district and AR stages “Li Bai” (李白) writing verses in space, and they also reference the Xi’an Museum (西安博物院) running the “Fengya Chang’an—Wang Wei Cultural Art Digital Immersive Exhibition” as part of a broader cultural branding agenda that uses immersive digital presentation and synthetic-character mediation to enhance visitor engagement.
Public Services And Governance: DeepSeek (深度求索/DeepSeek) is cited as being connected into Xinjiang’s government service systems, with a specific case in the Urumqi Economic and Technological Development Zone (头屯河区乌鲁木齐经济技术开发区) government service hall where a digital-human guidance agent is described as handling more than 770 service items, positioning the virtual presenter as an interface layer for navigation and service processing in e-government settings.
Education: Nanning Institute of Technology (南宁理工学院) is described as using digital-human lecturers to improve the efficiency of generating cross-language teaching resources, framing the synthetic lecturer as a scalable presenter/teacher proxy embedded into training and content-production workflows for application-oriented education.
Legal And Cybersecurity Outreach: Changsha Supercomputing Future Network Technology Co., Ltd. (长沙超算未来网络科技有限公司) is associated with a WeChat (微信) Mini Program labeled “Infinite Avatar AI Digital Human” (无限分身AI数字人) that appears in an official rectification/compliance notice, while Zouping City People’s Procuratorate (邹平市人民检察院) is described as producing a prosecutor digital-human public-warning item focused on verifying fee-collection claims in group chats, using a synthetic presenter to deliver procedural anti-fraud guidance to parents and schools.
Media And Heritage Digitization: Gansu Provincial Library (甘肃省图书馆) is described as using a “Wenlan Chongguang” (文澜重光) project to layer interactive technologies including digital humans onto the digitized Wensuge copy of Siku Quanshu (文溯阁《四库全书》) to provide an immersive online reading experience, and China Central Television (中央电视台) is referenced as a deployment context for a named digital human, “Digital Qian Xuesen” (数字钱学森), indicating broadcast media use of a synthetic historical figure presentation.
Entertainment And IP: Aida Era (艾达纪元) is described as a Hefei company oriented around “synthetic reality” and digital asset creation and is reported as having a seed round led by Chuangwei Capital (同创伟业), while Alpha Group (奥飞娱乐股份有限公司) is described as advancing an “AI+IP” framing in which digital humans are treated as activatable IP assets that connect character-driven content to toy and media ecosystems.
Public Education And Museum Experience: Yueyang City Museum (岳阳市博物馆) is described as building an offline physical AI digital-human docent system with interactive and performance-style explanation, and Kunming Information Port (昆明信息港) reporting references include a Kunming public-security digital human used in convenience services and Yunnan anti-drug outreach built around AI digital-human personas Qingqing (青青) and Lulu (鹭鹭), indicating a consistent pattern of synthetic-character presenters used to deliver public-information messaging through recognizable personas.
January 04 News
Enterprise And Service Terminals: Anhui Turing AI (安徽图灵人工智能科技有限责任公司) is described as advancing a movable, front-of-house “greeting” digital-human display terminal via a granted patent for a virtual-digital-human LCD device (CN223757230U) focused on improving connection stability in a mobile deployment, while Biyou Technology (必优科技) is described as patenting a document-to-roadshow workflow that outputs a synchronized presenter-style digital human whose lip movements, facial expressions, and body motions remain aligned to generated or provided audio, and Huawei Cloud (华为云) is described as featuring a digital-human livestream “proxy broadcasting” service as a marketplace case tied to local-business exposure/optimization work that integrates a large model identified as WIT大模型 to extract and foreground local business advantages.
Cultural And Tourism: The Wang Meng Literary Digital Experience Hall is described as using intelligent digital humans together with a cultural knowledge graph and big-data visualization to turn a corpus of writings, manuscripts, biographical trajectory materials, and Xinjiang cultural materials into interactive visitor experiences, and the tourism-IP examples describe AI digital humans embodied as historical figures—specifically an AI “Li Bai” at Gankeng Ancient Town with capabilities spanning classical poetry interaction through modern performance formats and incentive distribution based on engagement, and an “Xu Xiake” AI digital human embedded in a Kunming Daguan Pavilion on-site scripted experience that combines digital-human interaction with live enactment to make visitors active participants in narrative exploration rather than passive spectators.
Media, Marketing, And Content Creation: Jiufang Zhitou (九方智投) is described as packaging a “digital human” within a broader finance-oriented capability set that includes AIGC creation, sentiment/stance analysis, public-opinion analysis, and a financial knowledge graph, Shengtian Network (盛天网络) is described as offering an AI virtual-human podcast feature with selectable synthetic host avatars and personas for spoken content production, and Alibaba (阿里巴巴) together with Meituan (美团) is described in a video/animation capability context that emphasizes increasingly natural, fluent talking virtual characters, including an open-sourced avatar video generation model (LongCat-Video-Avatar) characterized as supporting emotional expression.
Live Broadcasting Infrastructure And Major Events: ByteDance (字节跳动), via Volcano Engine (火山引擎), is described as taking the role of exclusive AI cloud partner for the 2026 CCTV Spring Festival Gala, and Leo Group (利欧股份) is described as collaborating with ByteDance on virtual-human livestreaming through joint base-building and technology reuse for livestream applications.
Gaming And Large-Scale Participation: NetEase (网易) is described as applying “digital double” participation in a game context (逆水寒手游) to scale access to Dunhuang-themed cultural experiences, positioning digital selves as the mechanism by which extremely large audiences can engage beyond physical-site capacity constraints.
January 03 News
Cultural & Tourism: Yunnan Jiangwu Culture & Technology Co., Ltd. (云南讲武文化科技有限公司) is described as planning and executing an immersive, on-site roleplay script experience at Kunming’s Daguan Pavilion that uses the “Xu Xiake IP” and introduces an AI digital human version of Xu Xiake as an interactive guide; visitors can engage the digital human for interactive navigation and assistance while the experience combines roleplay and on-site games to encode local culture into a narrative format, with the project described as having undergone a multi-year development cycle before launch.
Enterprise & Recruitment: Baidu Online Network Technology (Beijing) Co., Ltd. (北京百度网讯科技有限公司) is described as applying for a patent for a “simulated interview” processing method based on a digital human, in which a digital human interviewer poses interview questions to job seekers, positioning the digital human as an automated interviewing interface implemented via a method, device, and electronic equipment.
Marketing & Live Commerce: Tekan Technology Co., Ltd. (Hangzhou) (杭州特看科技有限公司) is described as receiving patent authorization related to digital-human livestreaming, including a “real-time interaction” method for digital-human livestreams and a separate method for digital-human generation based on a deformation-network approach, both framed as system-level techniques intended to improve interactive performance and generation capability in digital-human broadcasting contexts; separately, 263 Network Communications (二六三) is described in an “intelligent marketing platform” context that emphasizes virtual digital-human capabilities and operational metrics such as sub–5-second virtual digital-human generation, high-accuracy speech-to-text, audience-intent recognition, marketing-effect prediction, and the use of industry-specific AI models (explicitly including finance compliance detection and education-domain models), positioning virtual digital humans as a core component of scalable livestream and marketing operations.
Technology & Interaction Methods: Alibaba (阿里巴巴) is described as addressing technical barriers in video animation for virtual characters via a method called “Knot Forcing,” portrayed as enabling digital/virtual characters to generate high-quality, time-consistent performance outputs driven by inputs such as audio, facial expressions, or pose signals to support more fluid, sustained conversational performance; in a separate patent-oriented interaction example, Guangzhou Weidao Technology Co., Ltd. (广州位道科技有限责任公司) is described as applying for a digital-human lip-alignment method intended to strengthen the interaction effect when users communicate with digital humans by improving mouth-shape alignment during communication.
January 02 News
Healthcare: Gushengtang (固生堂) announced an “AI avatar/digital twin” of renowned TCM expert Shi Qi (施杞), positioning it as the company’s first “national TCM master” digital avatar and framing it as part of a broader “data + scenario” approach to scale TCM AI services; the same material also describes a model where large numbers of doctor “AI avatars” are connected into a service network for initial AI triage followed by human physician verification or takeover, emphasizing that the virtual-being layer is meant to operate as a front-line clinical interaction and routing interface rather than a static video surrogate.
Entertainment: Digital Domain (数字王国) is described as applying long-running virtual human, VFX, and visualization capabilities to cultural/creative and “science-and-innovation” scenarios, and pairing that with HANBAO.AI (HANBAO.AI) to reduce technical bottlenecks in content creation so creators can focus on core creative work; separate items describe eventized uses of virtual beings at New Year’s shows, including a “cyber digital human” leading large-scale stage performances, and an elderly actor using a Ji Gong (济公) digital-human persona tied to a short-drama reservation push, indicating that recognizable characters and performer-linked synthetic personas are being used as promotional anchors for short-form screen entertainment.
Marketing: Huawei (华为) is referenced in a New Year’s city-event context through product-linked promotion (Huawei nova 15 / 华为nova15) embedded in local spectacle branding, while JD.com (京东) is cited for deploying an “AI digital human” version of a named procurement persona (“Caixiao Dongge” / 采销东哥) in retail categories to front campaigns and livestream-style commerce messaging; the same digest also includes “Huitaojin Digital Human” (慧淘金数字人) as a packaged solution used to produce leadership New Year greetings for organizational figures, explicitly framed as reducing production cost while improving communication efficiency and scenario innovation, and it ties that work to Zhongzhixinying (众智信赢) via a chairman-level client example, showing a pattern of templated, reusable spokesperson-like digital humans for institutional and executive communications.
Government And Public Services: A local prosecutorial office is described as producing a “digital prosecutor” (数字人检察官) to warn the public about “cloud fostering” fraud by using a synthetic presenter format to explain how fake farms and fabricated materials are used to lure victims, with the distribution/production pipeline noting Shandong Radio And TV ICT (山东广电信通) as technical support; a separate civic-consumption campaign note describes a digital human being used as an on-site policy announcer for provincial ice-and-snow consumption subsidies, reflecting a “virtual service figure” pattern where public-facing policy explanations and fraud-prevention education are delivered through consistent synthetic presenters rather than ad hoc human spokespeople.
Research And Technology: Westlake University (西湖大学) is linked to high-precision digital-human reconstruction research via a talk by Xiu Yuliang (修宇亮), describing three named lines of work—UP2You, ETCH, and Human3R—while emphasizing that digital-human reconstruction is trending toward becoming a base-model fine-tuning task rather than bespoke per-subject engineering; Yuanxi Lab (远兮实验室) is named as the lab associated with these projects, implying a technical stack centered on reconstruction fidelity and reusable model adaptation rather than one-off pipeline tuning.
Metaverse And Smart-City Platforms: Beijing Qingteng Technology (北京青藤科技股份有限公司) is described as obtaining a metaverse digital-human patent that uses deep learning to generate digital-human poses to improve interaction experience, positioning pose generation as a core enabling technology for real-time embodied avatars; Shiyou Technology (世优科技) is presented as an “AI digital human company” promoting a domestically adapted solution aimed at secure, intelligent upgrades for government exhibition halls, implying a deployment focus on controlled environments and compliance-driven infrastructure; Silu Visual (丝路视觉) is referenced in an investment-style context as having technology that supports digital-human production and development, with its digital-content toolchain framed as compatible with AIGC-era production; Wanda Information (万达信息) is tagged in a market-movement item with “virtual digital human” alongside smart-city and healthcare-related solution themes, suggesting it is being positioned as integrating virtual service figures into municipal/health IT application stacks rather than treating digital humans as standalone media artifacts.
Cultural And Arts: Beijing City Library (北京城市图书馆) is described as running a metaverse experience hall where three digital humans were “trained/fed,” with a Lu Xun (鲁迅) digital human noted as especially popular, illustrating a cultural-institution pattern where canonical historical figures are instantiated as interactive synthetic characters inside venue-based virtual spaces to increase audience engagement and interpretive participation.
January 01 News
Healthcare: Ant Health (蚂蚁健康) is presented through the “Ant A Fu” app (蚂蚁阿福App) as a deployment channel for “famous doctor AI avatars” (名医AI分身) that provide 24/7 health consultation, with the text describing quality improvements over iterative technical refinement (from low scores to many in the 70–80 range) and giving a concrete hospital-side example in which Beijing Friendship Hospital (首都医科大学附属北京友谊医院) deputy president Li Peng (李鹏) operates an AI avatar and reports high relevance to functional dyspepsia cases in gastroenterology clinics; the text also states that Peking University People’s Hospital (北大人民医院) partnered with Ant Health (蚂蚁健康) to establish a medical AI innovation joint research center and that more than 500 doctors have already launched AI avatars in the app, framed as a way to scale top specialists’ expertise to the public continuously.
Education: Aiweiwu (爱为舞) is described as delivering a “real-person-level” one-to-one AI tutor concept (真人级AI一对一导师) via a stack that explicitly combines a large model for “master-level” teaching methods with voice and digital human components to place the tutor into classroom-like scenarios, plus systems engineering to stabilize the end-to-end experience; separately, Shijiazhuang Post and Telecommunications Vocational College (石家庄邮电职业技术学院) is said to have filed a patent for an online education service system built around a virtual digital human (虚拟数字人) based on multimodal large models, and additional education-adjacent mentions position digital humans as part of “smart classroom” toolchains used to increase engagement in ideological-political teaching contexts.
Marketing: BlueFocus (蓝色光标) and Aofei Entertainment (奥飞娱乐) appear in a market-facing framing of a “virtual digital human” sector (虚拟数字人板块), with BlueFocus singled out among top gainers in a cited market snapshot and Aofei Entertainment linked to investor Q&A about whether multiple IP characters have corresponding virtual digital human images; Baidu (百度) is referenced via a statistic attributed to “Baidu company” that during the Double 11 shopping period a large share of merchants used digital humans to livestream, tying digital humans directly to high-volume commercial broadcasting, while Proya (珀莱雅) is referenced through commentary that digital human livestreaming still faces operational constraints around traffic and standardized operations, emphasizing adoption friction even when the use case is already normalized.
Enterprise And Public Services: Shengteng Innovation AI (Beijing) Co., Ltd. (北京昇腾创新人工智能有限公司) is described as having progressed from the “digital human track” into concrete industry outputs including a meteorological digital human (气象数字人) and a named digital persona “Xiao’an” (小安同学), while Ring Technology (环形科技) is described as having pioneered a digital-human reconstruction method constrained by motion-consistency (基于运动一致性约束的数字人重建) and being positioned as an AI-focused R&D/manufacturing/sales enterprise; Shiyou (Beijing) Technology Co., Ltd. (世优（北京）科技股份有限公司) is described as a full-stack digital human provider with a product referred to as the “Shiyou Bota AI digital human agent” (世优波塔AI数字人智能体) and a deployment approach that keeps data and systems on-premises for compliance, with concrete deployments spanning exhibition halls and government service settings and additional cultural-tourism style deployments described as digital human guides and multilingual explainers; DingTalk (钉钉) is named as a co-creation partner with Yunnan Construction Investment Group (云南建投) for an AI digital human named “Zhumeng” (筑梦) that appeared at a major expo context, and China Telecom Henan (中国电信河南公司) is described as applying digital humans within service systems such as navigation and online customer service, including a digital human customer-service capability claimed to recognize 200+ dialects to improve accessibility for older users; Industrial and Commercial Bank of China Xiong’an Branch (工商银行河北雄安分行) is referenced in connection with a “digital human” scenario inside a digital-currency smart enterprise-park context, framed as enabling very fast account opening and customer-facing convenience.
Cultural And Entertainment: MADverse Digital Art Space (MADverse 摩境数字艺术空间) is the venue for a strategic cooperation announcement involving ASTA (ASTA) and Zero One Culture (零壹文化), described as centered on a naked-eye 3D digital human performance IP (裸眼3D数字人演出IP) and positioned as a digital-art performance format rather than a purely informational assistant; Miaoyin Digital Technology (Miaoyin Digital Technology, 妙音数科) is described via a named digital persona “Lu Xingyue” (陆星月) that provides personalized explanations, multilingual service, and 24/7 intelligent customer support, extends into “restoring historical figure digital avatars” (还原历史人物数字分身) to enable cross-time dialogue with visitors, and integrates AIGC functions that can generate personalized travel notes in real time; Lan Zhi Mao (兰之猫) is named as a partner in a “digital culture and tourism” practice that uses an AI digital human capable of multilingual explanations and guided-tour style interactions, presented as a lightweight instruction-plus-card interaction pattern that triggers the assistant to respond in a specified language.