September 30 News
Healthcare: Shandong Digital Human Technology (山东数字人科技) presented anatomy-education systems based on digital-human modeling at multiple international and domestic medical-device and trade exhibitions, positioning digital humans as a repeatable training and demonstration medium for clinical education workflows. In Chengdu, a hospital blood-donation capsule integrated an AI virtual host based on the Chengdu Blood Center mascot “闷墩儿” to guide donors through the full donation process and generate commemorative videos, illustrating how an embodied virtual being can combine wayfinding, patient education, and post-event engagement inside a clinical service setting.
Marketing: Changzhuo Technology (畅卓科技) released a multi-pose AI avatar short-video system designed to preserve output quality under occlusion and varied capture conditions, pairing the tool with continuous support and operator training aimed at scaling synthetic-spokesperson production in short-form content pipelines. In destination promotion, Ningxiang’s municipal leadership used an AI digital human version of Party Secretary Zhang Zuolin in seamless alternation with the real person to introduce cultural and tourism assets at a development conference, showing a public-facing place-branding pattern where a named executive-like digital human is used as the presenter for civic marketing narratives.
Enterprise: Alibaba Cloud (阿里云) introduced the NuwaAI intelligent-agent digital human platform to reduce barriers to building and deploying digital humans, emphasizing minute-scale image generation and activation of dialogue capabilities so organizations can stand up branded virtual operators with less production overhead. Tencent (腾讯) enabled WeChat Official Accounts creators to build digital doubles supported by the Hunyuan model, with options such as knowledge-base support and prompt configuration to align a digital double’s output to a creator’s brand, publishing needs, and informational scope.
Public Sector And Civic Services: China Mobile Communications Group Henan (中国移动通信集团河南有限公司) and ZTE Corporation (中兴通讯股份有限公司) received recognition for a “5G New Calling AI Digital Human” deployment framed around accessibility and barrier-crossing service interactions inside enhanced calling experiences, indicating carrier-grade use of embodied digital humans as mediated service representatives. Harbin New Area’s “Beibei” AI government officer adopted a digital-human interface for public-facing information tasks, extending the same service-delivery logic into municipal contexts where a named virtual being is positioned as an official-facing representative for routine inquiries and guidance.
Events And Cultural Programming: Sixiang Wuxian (四象无限) and Feitian Yundong (飞天云动) surfaced applied virtual-being products at the Global Digital Trade Expo, with Feitian Yundong presenting the virtual digital human “Mo Xiaoyu” as a productized example of an event-ready synthetic character. Nanning Fengzhi Culture Communication (南宁风致文化传播有限公司) presented “Ju Xiaobai,” framed as a multilingual digital human that lowers story-to-screen barriers by auto-generating short dramas, which positions the virtual being as both a character layer and a production accelerator for serialized content. WorkBrain (工品一号科技) used its “1010 Digital Human Festival” and a V5.5 release to highlight an intelligent-agent direction for digital humans in event settings, aligning product narratives around more autonomous behavior and broader deployment readiness rather than one-off demonstrations.
Platforms And Social Companions: Soul (Soul App) (上海任意门科技有限公司) outlined plans to position virtual humans within its social platform as emotional-support companions and as bridges to human relationships, explicitly treating virtual beings as structured participants in social interaction rather than as decorative avatars. This plan emphasized raising the perceived “aliveness” of synthetic personas and formalizing supportive interaction patterns, which situates AI companions as a platform feature that is intended to shape user engagement, relationship formation, and ongoing retention through an embodied, named virtual-being layer.
Capture, Cloning, And Digital-Double Production: Hangzhou Second Life Technology (杭州第二人生科技有限公司) demonstrated an AI plus 3D-scanning booth that produces “human clone” digital doubles for deployment in games and metaverse spaces, paired with matching physical figurines, establishing a consumer-facing capture-to-runtime pipeline where a scanned individual becomes a reusable avatar asset. This same pipeline direction is echoed by enterprise and creator toolchains that treat digital doubles as a repeatable production unit, with the digital-human output designed for downstream insertion into interactive and media contexts rather than being limited to a single promotional clip or static portrait.
Standards And Legal Contexts: China’s National Radio and Television Administration (国家广播电视总局) approved “Digital Virtual Human Technical Requirements” (GY/T 411-2024), including a requirement for separate notice when processing a person’s voice or face, establishing compliance expectations that directly affect how digital-human systems are built, labeled, and deployed in media and service settings. Legal commentary referenced China’s first virtual digital human infringement case in parallel with broader accountability discussions in AI-assisted creation, framing practical fault lines around rights, responsibility allocation, and disclosure when real people’s likeness or voice is used to construct or operate a digital double within commercial or public deployments.
September 29 News
Entertainment: Baidu (百度) is cited with NOVA, a named “zero-latency” digital human positioned for live, synchronous presentation, indicating a use case where an embodied synthetic presenter is intended to operate in real time rather than as a pre-rendered host. ByteDance (字节跳动) is reflected through TikTok’s VTuber activity, where named VTuber personas such as chibidoki and catakemie use short-form video and stream promotion to run polls, mobilize fan engagement, and funnel audiences into live performances, illustrating an operational loop that combines avatar-based identity, audience management mechanics, and cross-format distribution from clips to live sessions.
Marketing: ChuangKeTu (创客兔) is described as publicising “Digital Human AI Talk 3.0,” with an emphasis on real-time script variable rewriting designed to stabilise livestream delivery for AI-hosted sessions, framing the digital human as a controllable on-air persona whose speech can be programmatically adjusted while maintaining a consistent presentation style. “Chufeng Premium” is presented as operationalising the virtual digital human “Chu Xiaofeng” as a brand ambassador to promote local products with high availability and risk-controlled messaging, indicating a deployment context focused on repeatable campaign execution and reduced variability compared with human talent. Complementing these examples, coverage on “AI selling face” describes market transactions in which AI digital-human personalities license their likeness or operate as virtual presenters, highlighting concrete commercial pathways for synthetic brand faces beyond one-off campaigns.
Gaming: Tencent (腾讯) is described as concentrating AI development within game pipelines while also promoting studio-linked virtual humans as audience touchpoints, tying character-based presences to both production capability and public-facing engagement. NetEase (网易) is framed similarly, with references that connect AI development priorities to game production and to virtual-human touchpoints that can maintain fan engagement outside the core game loop. miHoYo (米哈游) is linked to studio-associated virtual characters including “陆离_Louie,” “yoyo鹿鸣_Lumi,” and the user-reported “林离Olivia,” positioned as spokes-characters that can post content, sustain audience relationships, and potentially bridge into broader transmedia activations associated with game IP.
Enterprise: Alibaba Cloud (阿里云) is described as deploying the AI digital employee “Wan Xiaozhi” for enterprise workflows, indicating a virtual-being interface used as an operational role rather than an entertainment persona. Synthesia is referenced as providing a toolchain for creating corporate presenters and spokespersons as digital humans, aligning with an enterprise production model where avatar-led video output substitutes for on-camera talent and supports repeatable communications. NVIDIA appears as production infrastructure via Audio2Face, described as mapping voice to facial performance for avatars and digital humans to enable lip-sync and expression driving for real-time or recorded talking-head formats, and is also referenced as a strategic investor supporting the avatar tooling ecosystem, linking GPU-backed scaling and facial animation tooling to enterprise-grade deployment patterns.
Government And Public Services: Nanjing Jiangning District is described as integrating AI foundation models with operations tooling that includes a “Cockpit digital human,” deployed alongside intelligent work-order analysis and automated reporting to support municipal operations, placing a digital human inside an operational command context rather than as a public-facing mascot. Heilongjiang is described as deploying a public-service device that renders instant sign-language animations via a digital human across about 150 public venues, using an embodied synthetic character as an accessibility interface for inclusive information delivery. Shanghai is described as releasing a third batch of “Metaverse Application Scenarios” that explicitly includes digital humans as a core capability within a broader municipal stack spanning digital content, service platforms, 3D engines, and XR-related infrastructure, indicating that digital humans are being treated as a standardised capability within city-level platform planning.
Telecommunications: China Telecom (中国电信) is described through its Shanxi unit introducing a digital-human customer-service front end that fuses multiple AI capabilities with carrier service pipelines to deliver an interactive, animated support experience, positioning the digital human as the visible interface for service workflows. China Mobile (中国移动) is described through its Henan company deploying a “5G New Call AI Digital Human” integrated with next-generation calling, and the case is linked to recognition via an ICT China (2025) Case Award, indicating operator-grade validation for a digital-human calling experience. ZTE Corporation (中兴通讯) is named as the cooperating partner in the China Mobile deployment, tying the digital-human front end to telecom-network feature integration and carrier-scale rollout pathways.
Culture And Tourism: Guizhou is described as rolling out the “One Code to Travel Guizhou” 3.0 portal together with the cultural-tourism digital human “Huang Xiaoxi,” used as an on-screen AI tour guide that answers visitor questions and promotes local attractions, placing a digital human in a destination-information and promotion workflow. Chengde Mountain Resort is described as launching an AI digital-human guide for personalised cultural tours, and Ningxiang is described as featuring party secretary Zhang Zuolin as an AI digital human in a tourism promotion that explicitly switches between real and virtual selves to anchor city branding, showing the digital human used as a continuity device between human authority and synthetic representation. Shanghai Jiaotong University (上海交通大学) is described as presenting a digital human of Qian Xuesen during a stage performance, while other culture-tech deployments referenced include a digital human of Su Dongpo for interactive heritage interpretation, collectively positioning digital humans as narrators and ambassadors across tourism, heritage, and stage performance contexts.
Media And Public Events: CCTV Animation (央视动漫) is described as fielding the virtual character “Nezha” as a virtual delegate at a Hangzhou event, using recognisable IP as an embodied representative for public dialogue. Zhonghua Book Company (中华书局) is described as fielding “Su Dongpo,” Wasu Media (华数传媒) is described as fielding “Ai Jia,” NetEase Thunderfire (网易雷火) is described as fielding “DINO,” and Qiusuo Yingshe (求索影社) is described as fielding “Suo Suo,” with “Hangzhou Wenguang Xiaoyu” also referenced in the same event context, indicating a pattern in which media, publishing, and entertainment organisations deploy named synthetic characters as stand-ins for institutional participation. These cases treat virtual characters as delegates that carry brand or IP identity into civic and industry forums, rather than as standalone entertainment products.
Education And Training: A Guangzhou youth activity branded as a “Hello Audio Storytelling” is described as offering a hands-on “Digital human creation experience” in which student groups create digital-human assets within minutes, indicating entry-level tooling and curriculum framing around rapid avatar creation and media literacy. NetEase is also described in a separate training context as integrating large models and knowledge graphs with virtual digital humans for interactive learning and assessment, where the virtual human functions as a front-end mentor or guide and the back end personalises tasks and automates evaluation, indicating a closed-loop instructional design pairing synthetic instructors with data-driven curricula. The Yancheng Digital Audiovisual Industrial Base is described as exhibiting a digital-human production platform and real-time performance pipeline, mapping an institutional pathway from classroom creation exercises into professionalised production environments.
Ethical Contexts: Law-enforcement reporting described the use of AI deepfake pipelines to fabricate realistic female avatars for romance and investment scams, presenting a concrete misuse case where synthetic characters are operationalised as deceptive identities for social engineering and financial fraud. Separate reporting on AI companions describes consumer companions presented as on-screen digital avatars with facial expressions, body language, and lifelike vocal tone synchronised to chat, including commercial offerings framed as configurable “AI girlfriend” experiences with adaptive conversation and customisable traits, indicating a design pattern that emphasises affective presence and continuity. Tokenised “Virtual Companion” branding such as Neurosama is referenced as blending fandom, ownership signals, and access to the companion identity, connecting companion-style synthetic characters to financialised distribution models and raising practical concerns around disclosure, provenance, and the intensification of user attachment in avatar-mediated interactions.
September 28 News
Cultural And Tourism: Suizhou Yandi Cultural Tourism Area used a unified service avatar named “Yanyan” as a digital human guide delivered across mini-programs, web, and WeChat, positioning a single virtual being as a consistent interface for sightseeing assistance and related destination services, including route recommendation and visitor support. Hubei’s “Chufeng Premium” campaign used a virtual digital human named “Chu Xiaofeng” for regional product promotion, showing public-facing commerce and outreach being carried by a persistent synthetic character rather than rotating human presenters. Shandong Digital Culture Group (山东数字文化集团) deployed “Luyun” as a digital human interface built on the Yidian Tiancheng Digital Human Platform and a Qilu Culture Large Model to personalize cultural recommendations, while Shanghai Sports University (上海体育大学) used a Chen-style Taijiquan digital human to translate an intangible-heritage practice into a model-driven teaching and preservation tool, treating the digital human as both an instructional exemplar and a preservation artifact.
Retail And Live Commerce: JD.com (京东) deployed “Ta·Ta·Ta” as a universal digital human assistant that integrates chat, ordering, and device embedding, framing the avatar as both a service layer and an always-on operational presence within retail workflows. Zhejiang Turing Computing Research Institute (浙江图灵计算研究院) demonstrated multilingual avatar hosts for cross-border live commerce, aligning virtual presenters with language coverage, continuity, and staffing substitution in high-frequency livestream retail. Shandong harvest-festival operations used digital humans as substitute anchors to maintain non-stop streams, reinforcing the “always-on” virtual anchor pattern where synthetic presenters sustain continuous programming and customer interaction when human staffing is constrained.
Media And Events: Nanjing Xuanjia Network Technology Co., Ltd. (南京玄甲网络科技) used AI digital humans as stage-adjacent “actors” at expos, placing virtual beings alongside live programming as visible co-performers rather than back-office automation. Zhejiang Radio & TV Group (浙江广播电视集团) integrated digital humans into an end-to-end generative newsroom suite, positioning virtual anchors as front-of-house presenters within a broader production pipeline. China Mobile Guizhou Company (中国移动贵州公司) ran long-form AI digital human activities to seed adoption of avatar video formats and creator tooling at carrier scale, using events and programming to normalize virtual human production for broader audiences.
Gaming: NetEase (网易) described “AI doubles” in its mobile title “Nishuihan” that preserve “always-on” social bonds by allowing a player’s friends to interact with that player’s AI presence while the player is offline, embedding a persistent synthetic presence into the game’s social graph rather than limiting identity to active sessions. This positions the virtual being as a continuity mechanism for relationships and interaction, not only as a cosmetic avatar or scripted NPC, and it treats the AI double as a durable stand-in for the player’s persona within the game world.
Finance: JD Finance (京东金融) applied “Jing Xiaobei” as an AI digital-double interface for wealth management, enabling users to interact with an advisor’s synthetic counterpart for consistent guidance at scale. The case frames the virtual being as a professionalized service proxy that preserves a stable advisor persona while extending availability and throughput beyond what a human advisor alone can deliver.
Standards And Infrastructure: Baidu (百度) presented Wenxin-based, script-driven, multi-modal digital human control as a reference architecture for reliable avatar behavior grounded in large-model planning with coordinated speech and vision outputs, positioning the virtual-being layer as an engineered control problem rather than an improvised front end. China Internet Association (中国互联网协会) formalized taxonomies for media-type and service-type digital humans and solicited exemplary applications, connecting classification to procurement, compliance, and evaluation for enterprises deploying avatar staff for customer engagement, support, and sales. China Academy of Information and Communications Technology (中国信息通信研究院) and China Electronics Standardization Institute (中国电子技术标准化研究院) framed benchmarking attributes such as expressiveness and synchronization for virtual anchors to support vendor selection and performance optimization, while NVIDIA provided Audio2Face as a reference implementation for audio-driven facial animation that developers can embed into production pipelines to improve realism and synchronization in 3D virtual humans.
Healthcare And Training: Shandong Digital Human Technology Co., Ltd. (山东数字人科技) concentrated on medical simulation and education by pairing a high-definition virtual anatomy system (V2) with an intelligent CPR training system that ingests simulated vital signs for analytics, and by extending into clinical visualization via DIGIHUMANClinicalSystem. The combined portfolio treats digital humans as both representational anatomy assets and interactive procedural training counterparts, spanning classroom instruction, skills training, and scenario-driven clinical modeling.
September 27 News
Enterprise And Service Platforms: SenseTime (商汤科技) positioned its Digital Human platform as a customer-facing service layer that combines prosody-aware lip-sync with facial animation and gesture synthesis, framing the offering as a production-and-deployment stack for expressive virtual presenters and assistants in commercial service settings. Bangyan Technology (NuwaAI) (半言科技（女娲AI）) was described as providing an end-to-end “digital double” platform that covers external appearance generation, internal behavior construction described as a “soul,” and downstream deployment, with the stated aim of reducing creation and operational friction so enterprises and individuals can stand up agentic digital humans quickly. JD.com (京东) consolidated consumer-facing interactions around 他她它, described as a socialized digital human platform whose core agent 万能博士 supports question answering, small talk, and transactional tasks including food ordering, while also enabling users to create their own agents, positioning the system as a multi-service companion layer rather than a single-purpose bot.
Marketing And Commerce: JD.com (京东) extended the 他她它 ecosystem through a collaboration with 二元魂 (二元魂) that linked physical “trend toys” to a persistent virtual image, where squeezes and button presses update the virtual companion’s emotional state and memory, making the toy a control surface for an ongoing synthetic character relationship. Cloud Creature AI (云创兽 AI) (云创兽AI) was described as demonstrating commerce expansion through “digital twins” that allow one livestreamer to operate simultaneously in Chinese, Malay, and Vietnamese, presenting synthetic spokespeople as a way to parallelize multilingual audience reach without splitting creator time across separate sessions. Taobao (淘宝) used a 1:1 “Gaussian” digital human replica of an XR-related executive, Zhao Kun, as an on-stage co-announcer for a Vision Future Flagship Store opening, treating an executive avatar as a front-of-house retail launch asset in a brand event context.
Culture, Tourism, And Public Events: iFlytek (科大讯飞) appeared in a Guangxi ASEAN event deployment where two digital human emcees anchored proceedings on a multilingual stage, paired with seven-language real-time translation, emphasizing ceremony hosting and wayfinding as practical roles for digital humans in international convenings. Changsha municipal tourism programming used a digital human character to bridge live segments, while Hunan’s broadcast stack referenced XR virtual studios alongside a “Mango Digital Human,” presenting digital humans as narrators and connectors in modernized regional storytelling and tourism promotion. The China-ASEAN AI+Culture & Tourism competition was described as highlighting multilingual guidance and virtual digital humans as destination interfaces, reinforcing the pattern of digital docents and hosts integrated into event-stage and tourism-information workflows.
Education And Talent Workflows: Xi’an Aeronautical University (西安航空学院) adopted a digital human interviewer described as capable of sector-specific questions and real-time probing, using simulated human presence to standardize and scale candidate evaluation. Shanghai’s Huangpu District (上海市黄浦区) graduate fair deployed AI digital human interview stations for screening standardization, aligning with a broader pattern of virtual interviewers as throughput tools in hiring pipelines. Jiangxi educator training used a digital human to deliver child-safety instruction, and Hubei University of Technology (湖北工业大学) used a campus “digital human photo” installation for freshman onboarding, treating digital human presence as a scalable orientation and instruction interface across education settings.
Healthcare And Public Service: The Chengdu Blood Center (成都市血液中心) embedded its mascot Menduner as an AI virtual host inside a hospital “blood donation space capsule,” integrating the Aixin Cloud Platform (爱心云平台) to record the donor journey from registration through collection and to automatically produce a personalized short video at completion, while the virtual host continuously guided donors and sustained engagement in a clinical workflow. Guangfa Bank (广发银行) piloted a consumer-protection digital human experience zone that guided visitors through financial knowledge, positioning a digital human as an interactive educator for public-interest topics in a regulated service environment. These cases collectively presented virtual hosts and digital educators as structured interfaces for guided processes, with branded mascots and institutional deployments used to maintain continuity, explain procedures, and shape user experience in high-compliance contexts.
Telecommunications And Accessibility: Henan Mobile (China Mobile Henan) (河南移动（中国移动河南公司）) and ZTE (中兴通讯) operationalized an award-winning “5G New Call AI Digital Human” integrated into telephony, described as placing an AI presenter within the call layer to support “hearing and understanding,” with the deployment framed as barrier reduction through carrier-grade signaling and network integration rather than a standalone application. The same report cluster positioned digital humans as accessibility-oriented communication interfaces at provincial scale, emphasizing call-layer embodiment and service continuity as the distinguishing deployment context relative to event hosts and retail-facing digital spokespeople.
September 26 News
Healthcare: Sichuan Provincial Maternity and Child Healthcare Hospital (四川省妇幼保健院) operationalized a reviewed digital-human production line for medical popularization, indicating an institutional workflow for producing consistent digital-human health education content rather than ad hoc video creation. Gushengtang (固生堂) was reported as planning an AI “national medicine” twin intended for young doctors, positioning a named digital-twin concept as a training and practice support asset inside a traditional-medicine clinical context.
Enterprise and Retail: JD.com (京东) deployed a multi-role digital-human stack spanning the “Ta-Ta-Ta” universal digital human assistant, the upgraded “Joy,” the “Jingxi” shopping and life assistant, “JoyInside” for app control and task execution, and “Caishou Dongge” as a live-commerce presenter, combining advisory interaction, operational control, and on-camera synthetic presentation into a single commerce ecosystem. Hangzhou’s “Xizi Magic Box” sites (杭州“西子魔盒”) added a virtual digital-human runway and 3D virtual try-on to apparel districts as an experiential retail deployment that links digital-human presentation to apparel browsing and fitting workflows. 360 Smart Business (360智慧商业) launched the 360AI Digital Human Square with more than 200 roles and introduced N-World AI digital employees positioned for creative marketing and branded role-based deployments, emphasizing scale through a catalog of pre-defined personas rather than one-off builds. Changzhuo Technology (长卓科技) was described as providing a multi-pose short-video system designed to reduce production costs for brand avatars, framing digital humans as repeatable, low-friction commercial assets for short-form content pipelines, while separately reported livestream-commerce cases included a “Qingfou Digital Human” and a Luo Yonghao digital human presented as extending virtual beings into sales broadcasts with claims of more natural dialogue and motion.
Entertainment and Media: Huawei (华为) was associated with stage-grade, real-time digital-human performance through the “Yuri” digital human appearance at the Bund Conference, framing a conference-presenter role for a digital human in a live event setting alongside commentary about Huawei’s market entry. Kunming Information Port (昆明信息港) used a digital human anchor for legislative coverage, while Xining’s campus newscast avatar applied the same synthetic-presenter pattern to recurring school-news programming, showing adoption of digital humans as regular hosts rather than single-event demonstrations. Additional content-production examples included You Benchang’s micro-drama digital human and regional digital hosts in Shandong, extending the synthetic-presenter model into entertainment-adjacent short drama and local broadcast scenarios.
Culture and Tourism: Hanjingdi Yangling Museum (汉景帝阳陵博物院) deployed the “Shanshan” avatar for museum interpretation, aligning a named virtual being with onsite cultural explanation and visitor guidance. Wuhan East Lake (武汉东湖) was described as operating an Alibaba-backed metaverse experience that combines an “AI large model + digital human,” using a destination setting to embed digital humans inside immersive narrative and service flows rather than limiting them to static signage or audio guides. Qinghai introduced an AI tour-guide feature as a destination-facing deployment context for guided assistance, while Sichuan’s “Chen Zi’ang” digital host applied a named cultural figure as a digital-human presenter to convene events and deliver narrative framing in heritage-oriented programming.
Education: Harbin implemented a classroom storytelling avatar as an applied teaching aid in language-arts style instruction, using a synthetic character to deliver or scaffold narrative content in a school setting. Shanghai Jiao Tong University (上海交通大学) was reported as exposing students to a capture-to-model training pipeline, emphasizing how digital humans are built as well as how they are performed, and positioning the educational value in both technical production practice and downstream deployment literacy.
Industry and Workplace Services: Shandong Port (山东港口集团) implemented a conversational digital human to support finance personnel, framing a back-office assistant role in an operational enterprise environment rather than a customer-facing kiosk. Jiangxi Unicom (江西联通) deployed the holographic “Tongtong,” placing a named virtual being into a communications-provider context that emphasizes presence and presentation, while a Shanghai AI-promoter fair showcased staffing-sector digital humans and Qianfan-base releases included an “intelligent interview digital human,” collectively positioning virtual beings as tools for recruiting workflows, interviewing simulations, and demo-floor HR service scenarios.
Tooling and Creative Suites: Alibaba (阿里巴巴) was referenced through Tongyi digital-human demonstrations and through the East Lake metaverse case, linking large-model capabilities to digital-human interaction and cultural-scene deployments. SenseTime (商汤科技) was associated with the “Laozi” character as a culture-specific synthetic figure, indicating character-driven digital-human creation as part of a broader tooling and content suite rather than a single vertical application. GaiShi Technology (盖世科技) was cited for Shuode AI 6.0 as a multilingual digital-human capability, while an agentic-interaction stack attributed to Agaxi (阿加犀), Mingtu (名图), and Taren (塔人) was presented as supporting agentic interaction for virtual beings, together describing a production-and-interaction layer aimed at cross-lingual output, character creation, and scalable deployment of digital-human experiences.
Ethical Contexts: Zhu Songchun (朱松纯) was cited discussing intelligent agents integrating into human society, positioning digital humans within broader agent governance questions tied to social integration and responsibility boundaries. Separately, “AI resurrection” privacy warnings were reported as directly relevant to digital-human creation and use, framing risks around identity, consent, and data protection when virtual beings are built to reproduce a person’s likeness, voice, or behavior for persistent deployment across media, commerce, or institutional contexts.
September 25 News
Marketing And Live Commerce: JD.com (京东) deployed the digital human host Yanxi in live-commerce during Spring Festival trials and continued use during the “618” period, explicitly linking a synthetic presenter to measurable sales outcomes such as GMV, while Baidu (百度) supported a repeatable sales channel built around Luo Yonghao’s realistic digital-human stand-in that could appear consistently and deliver “one-image” AI product explanations without the scheduling and cost constraints of physical appearances. Changzhuo Technology provided a digital-human livestreaming system for cross-border and domestic e-commerce use, emphasizing multi-account operations, automated content generation, and image-driven product introductions as an efficiency product for brands, and Taobao (淘宝) incorporated digital-human livestreaming to promote local agricultural products during its Harvest Festival by organizing a platform-wide pool for virtual livestreaming. Nanning’s “Advertising+” initiative used AI digital-human livestreaming as part of an ASEAN-facing expansion and explicitly involved Chinese embassies and consulates as part of the outward market push, framing digital humans as scalable, policy-adjacent marketing infrastructure rather than isolated brand experiments.
Enterprise And Customer Service: Alibaba Cloud (阿里云) supported third-party avatar deployment through GladCube’s “AvaTwin,” described as a WAN-enabled AI avatar service delivered on Alibaba Cloud infrastructure, positioning cloud delivery as a practical scaling layer for enterprise-grade avatar services across locations. Chenjing Optoelectronics used digital humans as a traffic attraction and content carrier within a glasses-free 3D solution context, treating virtual presenters as a demonstrative layer to drive attention and convey content in a specialized display environment. The Beijing Economic and Technological Development Zone presented a “Live Digital Human” as a standardized module for corporate events and product launches, oriented toward automating enterprise content production for repeatable communications rather than one-off showcases.
Government And Public Services: Wuxi implemented an AI-agent matrix spanning 14 scenarios and 66 applications that included digital-human guides in service halls alongside mobile Q&A routing, showing a structured deployment pattern where virtual clerks support procedural navigation and service access across multiple channels. Xinbei District introduced the digital human “Xiaoxin” within a “Rule of Law Harbor” context to provide public guidance and process explanations and paired it with the “Su Jiefen” mini-program, while Jiangsu Province used “Digital Human Mentors” and original comics to deliver legal education as a combined media-and-avatar approach to civic instruction. Kaifu District’s “Xinghuo Xiaofu” was positioned as a virtual host for science-popularization activities to extend accessibility and online reach, and the Beijing Academy of Agriculture and Forestry’s “Agricultural Intelligence” platform enabled face-to-face consultations through “digital human interaction” within an agricultural knowledge-services setting, framing virtual humans as the interactive front end for expert systems rather than a decorative interface.
Culture, Tourism, Exhibitions, And On-Site Guidance: Wuhan-focused reporting on the convention and exhibition industry described “digital human interaction” as covering pre-trip planning, booking, and in-tour explanations with an explicit goal of 24-hour service at scenic spots and events, while the Hangzhou Digital Trade Fair and the Shanghai Advertising Expo integrated “interactive digital humans” into on-site guide and brand-interaction programs as a way to standardize visitor support and engagement on the floor. “Live Digital Humans” shown at exhibitions and parks were described as being used for live Q&A and hosting, and Qilu One Point (齐鲁一点) and Qilu Evening News (齐鲁晚报) described mass production of digital-human hosts and IP characters through the “One Point Tiancheng” platform, including provisioning multiple digital hosts for Jiaodong Online, linking venue-scale deployments to an upstream platform for rapid character creation and program integration.
Media And Entertainment: Zhejiang TV (浙江卫视) used the virtual host Gu Xiaoyu as a persistent presenter for Global Digital Trade Expo coverage, emphasizing continuity of event presentation via a stable virtual persona that can be reused across programming needs. You Benchang’s micro-drama applied digital-human techniques to recreate a younger Ji Gong, using a synthetic likeness workflow to address age and continuity constraints in screen storytelling, and municipal media examples described local programming anchored by the virtual “Xinghuo Xiaofu,” extending the same virtual identity across science-popularization and broadcast-like formats. Community media programming in Shenzhen described residents scripting and producing a “community news broadcast” using virtual image presenters, accompanied by practical guidance on rights and lip-sync, indicating an applied, production-oriented use of virtual presenters in localized public communication.
Platformized Creation Pipelines And Long-Form Digital Humans: Aliquark’s “Zhaodian” initiative and the “Long-form AI Digital Human” project associated with ByteDance (字节跳动) and Zhejiang University (浙江大学) were presented as demonstrations that long-form digital-human production and one-stop creative pipelines are becoming platform-based, with intended applicability across media, education, and marketing content production. This framing treats digital humans as a repeatable product category supported by standardized toolchains for generating extended-duration, humanlike presenter content, rather than as bespoke one-off character builds tied to a single show or campaign.
September 24 News
Healthcare: Medudy (KI avatars) deploys disclosed KI avatars as continuing medical education presenters for physicians, using avatar-mediated delivery to standardize instruction while keeping the modality explicit to avoid confusion about authorship and authority. Beijing City Library (北京城市图书馆) operates three digital humans that collectively supported more than 1.7 million voice-dialogue interactions, positioning digital humans as high-volume, spoken interfaces for civic information access that can support public-health-adjacent inclusion needs at scale, and it also uses “AI Lu Xun” as a synthetic author avatar inside a metaverse-oriented venue to guide reading promotion and structured learning pathways. Coverage of “expert agents” and doctor AI twins frames medical “AI twins” as assistive virtual beings that support explanation and triage workflows while not replacing licensed diagnostic decision-making, establishing a concrete boundary between supportive functions and clinical responsibility.
Entertainment: Tianzhou Culture (天舟文化) advances a virtual idol strategy that includes operating four virtual streamers through its subsidiary Youai, with named examples including The Fairy Next Door and Xiaoya, and it ties this virtual-being portfolio to adjacent development in VR games, cloud gaming, and cloud phones as part of a broader entertainment stack. Zhejiang Sci-Tech University (浙江理工大学) staged two virtual humans as ceremony hosts, demonstrating institutional event co-presentation where synthetic characters function as live-format anchors rather than background visuals. Kuaishou (快手) offers Keling AI digital human generation for creators, producing short creator videos up to 1080p and one minute, which positions a platform-native digital human tool as an entertainment workflow accelerator for rapid scripted or template-driven video output. WIMI Hologram Cloud (微美全息) describes always-on virtual-human operations for livestream formats and backstage automation, treating virtual beings as continuously running performance assets that sustain entertainment programming without the constraints of human scheduling.
Marketing: Baidu (百度) deploys a Huiboxing-powered rural livestreaming approach in Shandong that uses persistent virtual hosts to sustain after-hours sales, connecting virtual human continuity to measurable commerce endurance in low-traffic time windows. Taobao (淘宝) is described as adopting AI digital humans for its night-shift economy, using automated livestream sales to maintain retail presence during low-demand hours while keeping content output active. JingTeng (京腾) markets an AI marketing engine that includes smart montage generation and custom virtual digital humans, aligning rapid creative production with brand-controlled synthetic presenters for advertising and sales content. Agricultural Bank of China Hunan Branch (中国农业银行湖南省分行) runs a Digital Human IP Creative Competition that places a financial brand directly in virtual human IP development for customer engagement, treating digital human identity as a branded asset rather than a one-off campaign element.
Enterprise: Alibaba Cloud (阿里云) and Bangyan Technology (榜眼科技) collaborate on the NuwaAI intelligent-agent digital human platform, targeting creator and enterprise pipelines for virtual IP and digital doubles and positioning an agent-oriented digital human layer as a reusable production and operations substrate rather than a single application. DingTalk (钉钉) extends agentized spreadsheet integrations to include digital human generation and multimodal understanding, with Shandong Weiqiao Pioneering Group (山东魏桥创业集团) cited as deploying these capabilities for safety inspection workflows, which frames digital human tooling as a practical interface component inside industrial process governance. Volcengine (火山引擎), Jimeng AI (即梦AI), and Shanghai Film (上海电影) are associated with an AI Film Summit showcase that presents “Digital Human 1.5” as a production-grade presenter system, indicating an enterprise-ready packaging of digital human presentation into standardized stage and screen delivery.
Government And Public Services: Kunming Public Security (昆明市公安局) deploys “Chunxiao” as a large-model–powered digital human to deliver cybersecurity education and citizen guidance, using a named digital human as an institutional front end for public awareness and instruction. Mashang Consumer Finance (马上消费金融) uses “Ma Xiaoxiao” as a virtual counselor in campus programs to extend public financial literacy, treating the digital human as a repeatable educational presenter across student-facing contexts. Chongqing’s statistics bureau deployment is described as an AI digital human used for public communication and data outreach, which positions a synthetic presenter as an interface for explaining statistical work and distributing civic information in a consistent, scripted form. China’s National Cybersecurity Week demonstrations include AI face-swap, voice cloning, and digital human generation, using public exhibition contexts to show both capabilities and risks and to anchor responsible-use messaging around synthetic identity technologies.
Cultural: Hongguo Short Drama (红果短剧), Beijing Huanxidaoji Culture Media (北京欢喜道集文化传媒), and Shanghai Wenshuocheng Film (上海文说城影视) are linked through a micro-drama pipeline that applies digital de-aging within a virtual human workflow, making synthetic performance continuity central to short-form narrative production. Guangxi’s “Jiang Xiaomei” is presented as a localized digital human used for cultural spectacle and minority heritage presentation, while Fengxian Rongmei (奉贤融媒) uses a digital human to anchor civic festivals and media programming, keeping institutional cultural communication tied to an owned synthetic presenter identity. Nanning’s urban retail experiments deploy interactive city-guide digital humans in tourism conversion booths, and Changsha’s tourism conference integrates AI digital humans into destination showcases, positioning digital humans as event-scale place-branding instruments that combine performance, information delivery, and localized cultural signification.
September 23 News
Cultural: Beijing’s VR interactive narrative exhibition of The White Snake, supported by the Beijing Culture and Art Foundation, used volumetric capture, gesture recognition, and digital humans to re-stage Peking Opera inside an immersive, audience-interactive setting, positioning digital humans as both performance surrogates and guided pedagogical interfaces for traditional arts. Beijing City Library deployed three always-on digital humans—AI Luxun, Tutu, and 2122—accumulating more than 1.7 million voice interactions in a public cultural facility context and receiving the Vega Digital Awards Gold Award for Best Interaction, demonstrating sustained, large-scale operation of named virtual beings as front-of-house cultural guidance and engagement systems. In Shanxi, a cultural heritage exhibition combined naked-eye 3D with VR/AR/MR and an interactive digital human named Hua Mulan to deliver a historically framed experiential story, using the digital human as the visitor-facing narrative agent that mediates learning and museum/tourism interaction. Separately, Zhejiang University of Technology used two virtual humans in an opening ceremony to support staging and guidance functions in a campus event context, extending the same public-facing virtual-presenter pattern into institutional ceremony and public relations deployment.
Marketing: In China’s live commerce ecosystem, major platforms have been using virtual presenters and show hosts to maintain consistent product promotion and enable flexible content creation while reducing the operational burden associated with long-duration broadcasts, treating virtual beings as scalable on-camera talent rather than novelty effects. In fashion advertising tied to China-based production and distribution, campaigns increasingly use digital clones of real models as synthetic characters for rapid generation of diverse visual assets, leveraging AI and metaverse-adjacent pipelines to reproduce human expression without time-and-location constraints. TikTok is trialing AI-generated copy, synthetic voice, filters, and virtual human announcers as part of a distribution-layer approach that can automatically generate and read program progress, news, and promotional content, indicating a workflow where virtual beings become the delivery surface for algorithmically optimized messaging. These China-linked cases collectively frame virtual humans and digital doubles as repeatable, production-managed marketing labor embedded directly into commerce media operations rather than as one-off creative experiments.
Entertainment: Alibaba (阿里巴巴) was cited for the Wan2.2-animate model as part of an avatar and digital-human production stack aimed at higher-fidelity generation and reduction of uncanny-valley artifacts, indicating continued investment in models that directly target animated virtual human output for narrative media and creator workflows. Alongside this model-centric path, a practical creator pipeline was described through Blender-based workflows for building and animating lifelike 3D virtual humans for cinematic or social video outputs, showing how synthetic characters can be produced end-to-end with widely used tooling rather than proprietary studios. Virtual-performer ecosystems were also represented through ongoing VTuber activity on short-video platforms, where virtual performers release music, ship new avatar models, and update rigs as recurring production milestones that sustain audience engagement in platform-native distribution. These examples collectively describe entertainment-facing virtual beings as a combination of model-driven generation, commodity 3D production workflows, and continuous character upkeep for sustained performance and content cadence.
Enterprise: China-linked enterprise adoption was reflected through institutional deployments where digital humans are treated as operational front ends—maintaining persistent interaction loops, providing guided experiences, and acting as the user-facing layer for complex information delivery across public-service environments that resemble enterprise service desks in uptime requirements. The Beijing City Library’s long-run interaction volume with named digital humans demonstrates that virtual beings can function as durable service endpoints in real-world facilities, with performance measured through interaction counts and external evaluation outcomes. In parallel, event deployments such as Zhejiang University of Technology’s use of virtual humans for ceremony staging show how organizational communications can standardize presenter roles and scripted guidance through synthetic characters that are controllable, repeatable, and less dependent on human availability. Taken together, these cases illustrate enterprise-like operational patterns for virtual beings in China: persistent service delivery, event communications support, and interactive guidance designed for scale and reliability rather than novelty.
September 22 News
Healthcare: Medical avatars presented at the China-ASEAN Expo were described as clinician-adjacent, speech-enabled digital aides used for patient triage and multilingual intake communication, including ASEAN-language support intended to reduce friction in registration and initial questioning while standardizing patient-facing exchanges in multilingual clinical settings. iFlytek (科大讯飞) was also reported demonstrating on-site conversational virtual humans in exhibition contexts, reinforcing a pattern of public-facing medical or health-adjacent digital humans being deployed as workflow-facing communication layers rather than as clinician substitutes.
Government And Civic Services: Chongqing authorities operationalized Shancheng Xia as a network-security digital human for large-scale public education, while Tianjin Binhai New Area introduced Xiaobin for employment guidance and Hubei deployments included Xiaozhu in Zhushan plus municipal digital human avatars positioned for 24/7 civic responses. Zhangjiakou City Library in Hebei adopted a digital human narrator for resistance-history exhibits, and institutions tied these deployments to domain-specific messaging where the digital human functions as a persistent, accountable front-end for policy communication, guidance, and public-service information delivery rather than as open-ended companionship.
Commerce And Finance: Chinese live-commerce coverage described virtual humans being used as 24/7 livestream hosts to demonstrate and sell goods continuously, including examples framed around higher-priced items, with the operational emphasis on always-on availability beyond human presenter time limits. Digital Mali (数字马力) was described delivering AI digital humans across sectors including finance, internet services, 3C electronics, and restaurants, while Linhai Quyun Wanwei Information Technology (临海趣云万维信息技术) and a livestreaming complex in Wuhan were described configuring AI digital human anchors and unattended studios for round-the-clock retail, including linkage to Haier (海尔) via a national livestream center context. In financial risk-communication use cases, Bank of Communications Sichuan Branch and Bank of China Yunnan Branch were described deploying AI digital human presenters for anti-fraud and consumer-protection messaging, positioning the virtual presenter as a controlled channel for standardized prompts and risk education.
Education And Workforce Training: Guangdong organizations participating in a national vocational skills competition were described using AI digital humans as sparring partners and practice counterparts for WorldSkills-grade training, including language tutoring and coaching functions tied to institutional teams such as Guangzhou Industrial and Trade Technician College and project teams described as Shisaitong AI Intelligent. Tianjin Foreign Studies University was described embedding digital humans in ideological instruction and using virtual presenters for history content, while exhibition write-ups also emphasized visitor-facing conversational virtual humans as part of technology literacy and outreach at fairs and conventions, aligning training uses with broader public demonstration practices.
Cultural Heritage And Tourism: The Badaling Great Wall immersive theater in Beijing was described integrating XR, AI, digital humans, and game engines into a live cultural attraction, while Hong Kong venues were described combining digital humans with human docents and automated guides to orchestrate museum-style tours. A state-branded cultural avatar described as the Yellow River AI digital human was framed as conducting a cross-time conversation with the Nile to personify heritage narration and civilizational dialogue, and the Kunming South Asia–Southeast Asia Influencers Alliance emphasized interactive virtual digital human formats and AI-generated shorts for international storytelling, linking cultural communication goals with virtual-presenter interactivity.
Enterprise Productivity And Digital Twins: Tencent Meeting (腾讯会议) and Tencent Yuanbao (腾讯元宝) were described enabling users to delegate information-gathering meetings to a personal digital twin, framing a pattern of agentic attendance and summarization where a user-linked digital double participates in routine enterprise workflows. Jiuwei Health (九为健康) was described building clinician digital twins for traditional Chinese medicine experts using multimodal intake to emulate diagnostic workflows for scalable consultations, while market commentary also positioned cloud infrastructure as foundational to scaling virtual digital humans, citing Tencent (腾讯) and Huawei Cloud (华为云) as active producers in China’s broader metaverse-adjacent economy.
Media And Broadcast Presenters: A national TV host competition in China was described using XR virtual environments for immersive segments alongside conventional hosting tasks, indicating broadcast experimentation with virtual-production workflows around presenter performance. Mango Supermedia (芒果超媒) was described developing an in-house end-to-end virtual digital human stack intended for broadcast and education use, and iFlytek was described deploying dual digital human emcees as formal virtual hosts for the opening ceremony at CAEXPO in Nanning, alongside multilingual AI interpretation and mobile digital human demonstrations, positioning virtual hosts as stage-grade, institutionally branded presenters for large-scale events.
Tools And Production Pipelines In China: Keling AI (可灵AI) was described introducing digital human generation features that transform a single character image plus text or audio into up to 1080p video with precise lip-sync, controllable facial and body motion, and emotion-driven gesture modulation, explicitly positioning the tool as a lower-barrier pipeline for producing realistic, responsive digital characters across multiple styles. WiMi Hologram (微美全息) was described collaborating on DeepSeek (深度求索)-based digital human applications, and separate market commentary framed production and deployment scale expectations for virtual digital humans through 2026 with enterprise cloud as an enabling layer, situating toolchain evolution alongside infrastructure-driven rollout capacity rather than treating digital humans solely as standalone media artifacts.
Law And Governance: The Hohhot Xincheng District People’s Court in Inner Mongolia was described issuing a judgment framed as China’s first reported virtual digital human copyright case, clarifying infringement liability and compensable damages in a dispute over rights to a virtual digital human’s authored expression. Across other institution-bound deployments referenced alongside this legal marker, the governance pattern emphasized tying virtual presenters and digital humans to verified messaging, auditability, and domain-specific policy or risk content, aligning IP enforcement with operational controls where digital humans act as accountable communication endpoints in regulated or public-facing contexts.
September 21 News
Healthcare: Digihuman Technology (Digihuman Technology) is presented as deploying HD Digihuman Virtual Dissection Table System V2.0 for medical anatomy education, positioning a digital human interface as the instructional layer for guided dissection-style learning; in parallel, the Liaoning Medical Insurance AI Digital Human is described as a public-service “front desk” staffed by a synthetic agent for medical-insurance handling, and a separate case described via the Tibet Cyberspace Office frames an AI dietitian delivered as a conversational digital human providing nutrition counseling. In addition to these service-facing deployments, “AI Doctor Twin Support” is described as using a doctor’s always-available virtual “double” for health education, follow-up, immediate feedback, and personalized treatment guidance, emphasizing continuity of caregiver identity through a persistent personal virtual body rather than a rotating pool of staff.
Education: Changshui Xinyue (Changshui Xinyue) is described as deploying an AI psychological companion digital human, an AI smart playmate, and an AI psychological cabin oriented to student wellbeing through guided, embodied interaction patterns, while the Guangdong Provincial Vocational and Technical Education and Research Office and Guangzhou Mechanical and Electrical Technician College are described as using the Worldsmart AI system with AI digital humans for vocational skills training. Separately, Beijing Founder Electronics Co., Ltd. (Beijing Founder Electronics) and Shanxi University are described as training practitioners on digital-human media applications, connecting avatar-based production skills to institutional curricula and workforce preparation rather than one-off demonstrations.
Public Services: People’s Daily Online (Renmin Wang) is described as developing “Guilin Ling Er,” a hyper-realistic, 3D-modeled digital humanoid deployed as a tourism guide at the China-ASEAN Expo that provides real-time destination information in Guilin, recommends personalized itineraries, answers visitor questions, and connects audiences to virtual reality experiences via QR codes, placing the digital humanoid directly into museum, pavilion, and exhibition visitor flows. In employment services, Xining City Chengzhong District Employment Service Bureau and Qinghai Qingqian Culture and Technology Co., Ltd. (Qinghai Qingqian Culture and Technology) are described as running digital-human livestream job fairs alongside an AI job-seeking assistant and an AI entrepreneurship mentor to scale access and standardize guidance, aligning the “always-on” affordance of virtual beings with high-throughput public-facing service delivery.
Enterprise & Marketing: JOMOO (JOMOO) is described as integrating a voice-driven digital avatar powered by the DeepSeek large language model into a smart mirror-cabinet that functions as a data dashboard, framing an embodied digital human as the household-facing interface for IoT and data visualization rather than a disembodied assistant; in the same creator-to-enterprise continuum, KeLing AI (KeLing AI) is described as providing a “Ling Li” tool that generates short-form virtual presenters from images plus text or voice instructions and outputs 1080p, 48FPS video for enterprise scenarios including video assistance, micro-training, and product demonstrations. WiMi Hologram (WiMi Hologram) is described as partnering with DeepSeek (DeepSeek) to accelerate “Ling Li” development through an integrated production-and-operations solution ecosystem, and Industrial Bank Kunming Branch is described as distributing financial-literacy content via digital-human videos on branch screens and owned channels to standardize compliant outreach, tying virtual-presenter pipelines directly to regulated communications workflows.
Entertainment & Media: Nanjing companies and World Show are described as using a China–ASEAN Expo presence to generate ASEAN-language digital humans on site and promote AI micro short dramas, including a Cambodia-focused production, positioning digital humans as localized actors, narrators, and hosts for cross-border short-form media; this aligns with a separate description of the Nanjing AI micro short drama initiative in which digital humans function as scripted on-screen talent rather than generic explainers. In professional broadcasting, “AI presenters” used as synthetic on-screen hosts within YouTube content and “Talking Avatars for Broadcasters” are described as production workflows that accelerate script-to-screen pipelines for news and programming, emphasizing operational adoption of virtual anchors and automated delivery frameworks in media production.
Legal & Ethical Contexts: The South Korean court ruling involving PLAVE is described as recognizing that reputation can extend from a real performer to a linked virtual performer when the connection is publicly known, enabling damages claims for insults directed at the virtual image and shaping moderation and reputation-protection expectations for virtual-idol deployments. Separately, the refusal by Pope Leo XIV to permit an AI avatar of himself is described as foregrounding identity, authenticity, and spiritual concerns that can constrain institutional deployment of synthetic personae even where technical creation is feasible, illustrating how public-figure governance decisions can set practical boundaries for virtual-being use in religious and cultural settings.
September 20 News
Healthcare: Ant Group (蚂蚁集团) was reported as advancing virtual clinician use through its AQ initiative, positioned as an AI “doctor twin” and decision-support layer aimed at improving access and physician scalability, with distribution anchored through integration into Alipay and links to nationwide providers; in parallel, Zhizhen Technology (智诊科技) was described as developing WiseDiag for physician modeling and community education, including rural-oriented deployment narratives such as building elder health records and learning from AI suggestions during care workflows. Separately, at the 22nd China–ASEAN Expo in Nanning, a multilingual “ASEAN-language medical avatar” was demonstrated as a practical bridge for doctor–patient dialogue across languages in clinical settings, and the same digest referenced a “High-Definition Digital-Human Virtual Anatomy Platform System V2.0” recognized at the Canton Fair Design Innovation Awards as an anatomically detailed, manipulable virtual-body system intended for lecture, lab, and museum-style instruction where ethics-compliant visualization replaces cadaver-dependent demonstrations.
Public Services And Finance: Mashang Consumer Finance Co., Ltd. (马上消费金融股份有限公司) was described as deploying a consumer-protection digital human during Financial Education Publicity Week activities in Chongqing to raise awareness of customer rights and safeguards, while China Guangfa Bank (广发银行) was reported as introducing a “financial knowledge science officer” digital human through its Kunming branch for outreach and education. Bank of Communications (交通银行) Hainan Branch was reported as using a cybersecurity-themed livestream fronted by its digital human employee “Jiang Ge” to deliver compliance-sensitive guidance at scale, and China Unicom (中国联通) Shandong was described as running visitor-facing “digital-human makeover” experiences at a provincial science venue that automatically generated personalized transformation videos, illustrating one-to-many avatar content generation as a public-engagement format rather than a purely internal communications tool.
Governance And Civic Administration: China Telecom (中国电信) was reported as operating multiple service automations including “Xiao Ying” in Guangxi for 24/7 government and livelihood Q&A, alongside a plan by Gansu Telecom to upgrade the 10000 hotline with video-capable agents and phased introduction of AI digital humans plus AR-style annotation to strengthen human–AI collaborative service. Daokeyun Tech (道客云科技) was described as pushing an “AI-for-governance” government metaverse solution built around digital humans to standardize omnichannel, scenario-based public-service interactions, and the same set of reports also framed a “metaverse for governance” pattern that includes multilingual virtual digital humans such as “Xiao Tong” for real-time policy Q&A, presented as a template for 24/7 civil services and trade-facilitation style interactions.
Culture, Museums, And Tourism: People’s Daily Online (人民网) was reported as presenting “Guilin Ling’er” as an ultra-high-resolution culture-and-tourism digital human used for destination storytelling and events, and Guangxi deployments were also described as using a holographic digital human “Jiang Xiaomei” with voice, natural-language understanding, and rendering integrated for audience assistance in live settings, alongside expo-stage digital human emcees dressed in regionally specific attire to manage on-stage control and agenda flow. Shengshu Technology (生数科技) was described as using Vidu AI to produce what was characterized as a first domestic AI culture-tourism film for Beijing’s Central Axis, using virtual hosts to guide viewers through historically significant sites, and Hebei museums and libraries were reported as using digital human explainers in cloud exhibitions to narrate anti-Japanese War history via virtual docents, indicating sustained institutional use of synthetic presenters for scripted historical interpretation rather than only for marketing clips. Zhongnan Cartoon (中南卡通) was described as deploying an AI holographic interaction pod that places a virtual character on site for real-time, gesture-responsive conversation, enabling museum-style encounters without requiring an in-person human presenter.
Telecom, Venues, And Transport Experiences: Qingdao Metro (青岛地铁) was reported as incorporating digital human elements into an award-recognized brand story titled “Hello, The AI Era of Urban Rail Transit,” using virtual presenter sequences to communicate rail operations, passenger services, and safety messaging in a controlled narrative format. Shenzhen Guangdian Xinyi Technology (深圳广电信义科技有限公司) was reported as demonstrating a smart-venue operations cloud integrating digital human guides and AR navigation, positioned as an accessibility and inclusion-oriented visitor-service layer for large public events, and the China–ASEAN Expo coverage also described “digital-intelligent” customer service for transport scenarios with instant translation and avatar-mediated interfaces to support multilingual travel queries and itinerary adjustments in real time.
Education And Training: Chengdu University (成都大学) was reported as launching an 80-episode digital human micro-course series on the “Xuexi Qiangguo” platform, with delivery framed as foundational digital-literacy micro-lessons developed by faculty in film and animation and produced in collaboration with Sichuan Communication University (四川传媒学院). Additional vocational and compliance training deployments were described as using digital human narrators for concise, scenario-based disciplinary education in state-owned enterprise integrity training and as AI digital-human teaching components at national skills competition venues, positioning the virtual presenter as a standardized instructor interface for repeatable curricula rather than a one-off promotional spokesperson.
Industry Tools And Production Pipelines: Kuaishou (快手) was reported as introducing Keling AI (Ke Ling AI) with a digital human generator that produces up to one-minute talking-avatar videos at 1080p/48 FPS from a single character image plus text or audio, targeting short-form explainer, course, and news formats with minimal setup, while Wondershare (万兴科技) was described as augmenting Filmora with AI features including virtual humans as part of broader creator-economy editing workflows. Hefei Tongxu Smart Technology Co., Ltd. (合肥同旭智能科技有限公司) was reported as combining display hardware with proprietary AI algorithms to improve real-time facial modeling and lip synchronization for production-grade digital humans, and an expo-focused “Yellow River and Nile” digital human co-creation video was described as using synthetic narration for cross-civilization public diplomacy, indicating that the same production stack is being applied to both creator tools and state-adjacent communication outputs when virtual presenters are preferred for consistency and multilingual reach.
Legal And Rights Enforcement: The Hohhot Xincheng District People’s Court in Inner Mongolia was reported as issuing what was described as the region’s first copyright-infringement decision involving a “virtual digital human,” recognizing a virtual concert as an audiovisual work and addressing unauthorized rebroadcast, while earlier Hangzhou Internet Court jurisprudence was cited as groundwork for disputes around virtual-human likeness and performance rights. The legal coverage summarized an emerging operational implication across use cases—virtual streamers, in-game characters, and digital-human customer-service agents can embody protectable creative expression—so operators are expected to manage licensing, anti-piracy monitoring, and compliance for avatar assets used in live commerce, exhibitions, and service delivery where recordings and re-uploads are routine.
Ethical Contexts: The digest grouped multiple governance and mental-health concerns around increasingly video-realistic virtual people encountered outside clear disclosure, emphasizing trust, authenticity, and the risk of persuasive deployment when a synthetic persona is presented as a credible social actor. It also summarized research and commentary on AI companions, including an MIT-linked analysis of “AI girlfriends,” and clinical reporting on so-called “AI psychosis” that noted limited scientific grounding while still documenting cases of distress associated with intensive companion-style engagement, framing these issues as practical policy and duty-of-care concerns that intersect with companion design, disclosure practices, and the moderation of high-frequency, relationship-like interactions.
September 19 News
Healthcare: Zhizhen Technology (智臻科技) deployed Haoban AI to construct AI “clones” of medical experts from major hospitals including Ruijin Hospital, Union Hospital, and Run Run Shaw Hospital, combining a proactive AI engine with individual health records and historical information so the expert-modeled companion can provide personalized, continuous guidance and follow-up-style engagement within clinical information workflows. Liaoning’s “Yibao’er (Liao Xiaobao)” medical-insurance AI digital human extended the same digital-human pattern into public-facing health-adjacent administration by providing rapid, accurate policy responses in a government service context, demonstrating how a named digital persona is used as a consistent front end for high-volume question answering where reliability and standardization are primary deployment goals.
Government Services: Qinghai’s Smart Road deployment at Qinghai Lake embedded a digital-human broadcast layer into roadway guidance so trip information can be delivered in a more personalized, continuously available manner without relying on venue-specific staffing or fixed-site programming, showing a transportation-context avatar used as an operational communications interface. Jiangsu’s campus cybersecurity program used a neutral avatar co-reader to standardize safety messaging at scale, and Chengdu’s cybersecurity-week exhibitions similarly let the public trial digital-human dialogue and embodied agents in a large event setting, linking administrative education campaigns to interactive digital-human experiences rather than static materials.
Culture: Shaanxi’s coordinated digitization across the Shaanxi Archaeological Museum, Xi’an Beilin Museum, the Hanjing Emperor Yangling Museum, the Xi’an City Wall, and the Shaanxi Provincial Culture Center treated digital human docents as standard front-of-house infrastructure for tailored guidance, and the Hanjing Emperor Yangling Museum’s “Shanshan” IP explicitly extended beyond interpretation into derivative product development, connecting a named avatar identity to revenue-linked cultural merchandising. Fujian’s “Hehe” 3D digital human for Quanzhou was reported in terms of high-volume service delivery and satisfaction, while Shanxi’s “Jin Yiyi,” an interactive “digital poet,” and a holographic “Mulan” demonstrate a portfolio approach that mixes on-site narration with networked promotion using multiple avatar formats under distinct named characters.
Tourism: Hunan’s Hongjiang district used an event digital human as an anchor within a broader immersive slate that also included the VR title “Biaoju Hubao” and the MR title “Caishen Sibao,” indicating a deployment pattern where a named digital presenter is integrated alongside immersive media products to structure visitor-facing content and programming. At the 22nd China-ASEAN Expo, World Show (世优科技) operationalized virtual beings for on-site multilingual generation, AI live streaming, cross-border digital marketing, and influencer-economy tooling, and Tianxiaxiu (天下秀) separately demonstrated on-site generation of multilingual ASEAN digital humans using its Linggan Island platform paired with TOPKLOUT (Kelaorui) analytics, establishing a concrete “generation plus measurement” stack for cross-border tourism-adjacent promotion and event marketing.
Marketing: Yantang Dairy (燕塘乳业) deployed “Cow Tangtang” as an always-on virtual mascot for brand identity, illustrating a persistent synthetic character used as a continuity layer for marketing presence rather than a one-off campaign asset. In the same China-ASEAN Expo context, World Show’s multilingual avatar generation and AI live-stream tooling aligned with Tianxiaxiu’s Linggan Island plus TOPKLOUT (Kelaorui) pairing for cross-border brand communications, and Baidu Youxuan (百度优选) served as the venue for Luo Yonghao’s scheduled digital-human livestreams, indicating preproduced avatar retail sessions that decouple talent availability from sales-window timing while keeping a recognizable persona as the sales-facing interface.
Education: Shandong Digital Person Technology Co., Ltd. (山东数字人科技有限公司) advanced professional medical and anatomy education with the “High-Definition Digital Human Virtual Anatomy Table System V2.0,” which was recognized with the Canton Fair CF Gold Award, positioning a named digital-human system as a formal teaching instrument rather than a general-purpose avatar. At the other end of the education spectrum, Hunan’s “Qixiaoyue” avatar in the Qiyang library provided reading guidance and desk-level information services for young patrons, demonstrating a child-facing informational digital human embedded in public cultural infrastructure to support routine queries and literacy-oriented navigation.
Media And Entertainment Production: Hengxin Oriental (恒信东方) created a hyper-realistic digital double of Jin Yong within an entertainment pipeline, treating a named cultural figure as a digitally embodied character asset for production use, while the feature film “Emei” publicized AI digital-human usage inside a sci-fi wuxia pipeline in the context of Chengdu film-industry signings, indicating adoption of digital humans as production elements in commercial film development. Keling AI (可灵AI) introduced a toolchain that converts a single character image plus text or audio into a 1080p/48-fps, one-minute digital-human performance clip, directly targeting fast turnaround needs for short-form media by turning minimal inputs into a complete, time-bounded synthetic performance.
Enterprise And Public-Facing Recommenders: Shandong Digital Culture Group (山东数字文化集团) operated the “Luyun” 3D hyper-real recommender on the Yidiantiancheng digital-human platform and anchored it with the Qilu Culture Large Model V2.0, which was rated Level-5 by the China Academy of Information and Communications Technology (中国信息通信研究院), tying a named digital-human recommender to an explicitly identified model backbone and an external evaluation claim. Within administrative procurement and vendor signaling, World Show’s appearance as an event-and-marketing provider sits alongside a separate recognition that positioned World Show as a preferred provider for government digital humans, illustrating how digital-human deployments are framed both as operational interfaces and as supplier-classified products within public-sector ecosystems.
Legal And Governance Contexts: The Inner Mongolia ruling treated virtual idol concerts as protectable audiovisual works and held unauthorized retransmission to be infringement, establishing that avatar-fronted performances can be defended as formal audiovisual outputs rather than informal online content. Separately, a U.S. action by Disney against a Chinese generative-AI company highlighted rights boundaries around character usage and AI-rendered character embodiments, while domestic governance debates also encompassed consumer-protection scrutiny of “AI companions” described as causing harms and triggering calls for safeguards, collectively defining enforcement and safety concerns that track synthetic characters across performance, brand, and companion-style interaction contexts.
September 18 News
Government And Public Services: Jiangsu’s information-communications sector deployed “Su Xiaotong” as an anti-fraud digital human for public outreach and feature education across campuses and partner venues, while Kunming Public Security introduced the dialect-capable AI digital human “Spring Dawn” to handle public-service interaction so residents could reduce in-person trips; additional local-government digital humans were launched in Xining and Mianyang to streamline administrative services, and Shanxi introduced the 3D virtual digital humans Jin Xiaoan and Tong Xiaoxuan as cybersecurity ambassadors to deliver public-interest messaging at scale. Alongside these citizen-service deployments, Baidu (百度) launched a legal-services initiative that uses lawyer AI avatars to provide localized guidance and action pathways, and Cryptography Technology (Jilin) Co., Ltd. created an AI Digital Human Alliance around an AI+NFC short-video engine positioned to accelerate digital-human-driven content creation, while iiMedia connected these rollouts to technical maturation constraints such as earlier motion-capture and rendering limits that affected facial expressiveness and latency.
Healthcare: Beijing Tsinghua Changgung Hospital used the holographic digital human “Rongrong” for clinical education and patient-facing explanation, paired with an “AI hepatobiliary super-assistant” for doctors to support decision workflows inside hospital settings. In parallel, AQ was described as aggregating consumer health functions, connecting to nationwide providers, and exposing more than 300 famous-doctor AI personas, while a Guangzhou enterprise’s “Yundie ‘Xingzhi’” large model was presented as generating AI teaching assistants and professional “digital doubles” for doctors to transfer diagnostic knowledge, framing these virtual experts around triage support, report interpretation, and medical education under regulated guardrails.
Education And Talent: Mediazen positioned avatar instructors as a concrete delivery format by combining an AI English Library and digital-human lectures with VR experiences for immersive language learning, and this educational use case was linked to enabling layers such as Aiweiwu’s low-latency voice-to-animation framework and Caixun Co., Ltd.’s interactive digital-human stack for synchronizing speech, expressions, and multilingual content in responsive teaching personas. In formal schooling and workforce preparation, Nanjing Keliwa Middle School used a “digital human study companion” led by teacher Tian Xin for synchronous English-learning interaction, while the Changping Cultural Creativity Competition added an audiovisual AIGC track covering digital human development; the Jiangxi Digital (Industrial Software) Talent Training and Innovation Base—co-built by the Jiangxi Provincial Department of Education, the Ganjiang New Area Administrative Committee, and Huawei Technologies Co., Ltd. (华为技术有限公司)—was described as building a skills pipeline for digital-human content and systems, and Founder Electronics collaborated with Shanxi University on an applied AIGC camp that included digital human application training.
Commerce And Marketing: TopSocial demonstrated multilingual virtual humans, AI live streaming, and cross-border digital marketing at the China–ASEAN Expo as a mechanism for influencer commerce and brand expansion, and Luo Yonghao’s Make Friends team scheduled digital-human livestreams on Baidu Youxuan to run retail programming with synthetic hosts. In the same exhibition context, iFLYTEK (科大讯飞) presented a photo-to-avatar virtual-human interaction platform designed for low-effort creation of personal digital presenters used in multilingual engagement at exhibitions, and NuoYun’s partnership with Tencent Cloud (腾讯云) was characterized as embedding AI virtual humans alongside blockchain traceability into private-domain live commerce, aligning synthetic hosts with verifiable commerce operations.
Enterprise And Industrial Platforms: Baidu (百度) expanded its digital-human generation and vision-model platforms with the stated goal of reducing enterprise costs and improving deployment efficiency, while Caixun Co., Ltd. described a productized stack covering real-time video generation, scene transitions, and personalization to support operational virtual presenters. These platform moves were situated as consolidation across authoring, orchestration, and runtime layers that lower barriers for deploying virtual humans across enterprise workflows, and Jiangxi Guanying Intelligent Technology Co., Ltd. was described as integrating digital humans and digital twins into industrial AI terminals to support cluster digitization, with Nanjing’s Yuhuatai District and Kunshan positioning digital-human companies within software and industrial parks to connect supply chains.
Finance And Consumer Education: Ping An Bank (平安银行) received authorization for a patent on virtual-image construction methods framed as formalizing avatar-creation approaches for financial-services scenarios, indicating continued investment in controlled virtual personas for regulated customer-facing contexts. China Construction Bank (中国建设银行), via its Guangzhou Yuexiu Branch, deployed the Far Eastern Digital People for consumer financial education, pairing institutional messaging with an explicitly branded digital-human presence.
Culture, Tourism, And Exhibition: Tianjin attractions adopted AI digital humans as smart customer-service agents to support visitors, and Hubei’s Culture and Tourism Industry Expo showcased digital humans as museum interpreters, including deployments at Hubei Provincial Museum; the World Manufacturing Convention in Hefei also introduced a “digital human + docent + automation” mode for exhibit explanation. Shandong Digital Culture Group presented Lu Yun as a 3D hyper-realistic digital human distributed on the Yidiantiancheng digital human platform for cultural content circulation, and the Qinghai Lake Smart Highway system used an AI digital human to publish time-sensitive travel information such as hotel bookings and parking saturation via roadside displays, with an outcome claim of reducing average information-acquisition time from 15 minutes to 30 seconds by integrating virtual-human updates into physical tourism infrastructure.
Media And Broadcasting: Broadcasters in Anhui normalized the use of virtual anchors, including a named example “Xiaojia,” as part of routine program production, indicating continued use of synthetic on-air characters for scripted delivery within broadcast workflows. In parallel, Zhejiang’s Global Digital Trade Expo announced an “AI digital human IP steward” for rights management and also promoted cross-border live streaming with humanoid presenters, framing digital humans not only as presenters but as IP-managed assets within trade, licensing, and exhibition ecosystems.
September 17 News
Marketing And E-commerce: A.S. Watson Group deployed the virtual spokesperson “屈晨曦 (Qu Chenxi)” to act as a branded front-facing figure for consumer engagement, while Florasis (花西子) partnered with a dedicated “Florasis virtual person” to deliver brand messaging in beauty retail contexts. China Merchants Bank (招商银行) rolled out the virtual service persona “AI小招 (AI Xiao Zhao)” as a branded touchpoint for customer-facing communication in financial services, and WEIHUA Group (卫华集团) integrated a 24/7 marketing digital human into its industrial AI platform to sustain continuous customer outreach. Baidu (百度) used Baidu Select to host two “digital human” livestreams operated by Luo Yonghao’s team to broaden content and transactions, and Kuaishou (快手) promoted “digital human” livestreaming in apparel to reduce labor and time-slot costs while improving repeatability of commerce content; Monster Lab (怪物实验室) was described as incorporating “digital humans” into routine in-store livestreams as part of a platform-to-merchant implementation chain.
Enterprise And Industry: China Merchants Group (招商局集团) deepened cooperation with Baidu to apply large-model agents and digital humans or “digital employees” to business scenarios framed as having measurable outcomes, positioning digital humans as operational interfaces alongside agent workflows. Jiangsu’s Virtual Reality Industry Innovation Center was described as building regional capacity around XR, large models, 3D engines, and virtual digital humans to support enterprise-grade creation and deployment pipelines, and Henan Province’s action-plan framing proposed using “digital people” overlaid onto 3D product models to provide real-time consultation and personalized configuration for sales-front and solution-design workflows that connect pre-sales to customized production.
Government And Public Services: China Telecom (中国电信) in Nanning was described as establishing “digital humans” as community service touchpoints in smart communities serving residential users, while Ant Digital Technologies (蚂蚁数科) was described as providing a “Digital Human Employment Kiosk” intended to integrate public employment management and matching for governments, businesses, and job seekers via close-range service delivery. Baidu was also described as fielding a legal intelligent agent that connects residents through a QR code to a lawyer avatar for one-to-one public-service consultation, and multiple local deployments were cited including Tianjin’s Beichen District using a “Digital Human” named “Xiaochen” as a government consultation front desk, Gansu’s “Lan Xiaoxin” co-hosting large events with human presenters while serving as an external information channel, Qingyang using on-call “digital humans” to provide continuous consulting support to businesses, and Guangzhou’s sports-event platform integrating “AI Digital Humans” into city-level command as presentation and interaction interfaces.
Cultural Tourism And Public Exhibitions: China–ASEAN culture-tourism initiatives were described as featuring multilingual smart guides and virtual digital humans paired with smart scenic-area management aligned to ASEAN visitor contexts, treating digital humans as culturally adapted presenters for heritage and travel storytelling. Additional China-focused examples described “digital humans” guiding tours of the Qin Mausoleum’s underground palace through immersive heritage explanations, Tianjin deploying intangible-cultural-heritage “digital humans” for knowledge expression and interactive dissemination, the Central Plains Cultural Expo combining “virtual digital humans” with AIGC installations to increase visitor interaction, the Guangzhou Paralympic Games using “digital human customer service” and AR guides to improve accessibility and visitor engagement, and Henan Province using “digital humans” to promote “Henan-style romance” as part of regional cultural narrative-building.
Healthcare And Social Support: Haodafu Online (好大夫在线) was described as building more than 300 physician AI “twins” as doctor avatars that interact continuously with patients, extending medical guidance across a network described as having hundreds of thousands of online practitioners. Public-sector deployments were described in which Liaoning Medical Insurance used “AI digital humans” to provide policy and application guidance to insured residents, multiple hospitals promoted AI-assisted Traditional Chinese Medicine diagnosis and treatment to improve triage and advice accessibility, and Shenyang’s “AI Fate Engine” combined “digital humans” with social and companionship services for people with disabilities, presented as linking medical care, medical insurance, and community care into a coordinated service model.
Education And Learning: Education deployments were described in which a “digital human learning partner” participates in classroom instruction and homework discussion, providing instant feedback on learning tasks and creative design work while shifting interaction from one-way playback to ongoing two-way engagement. This educational pattern was presented as a concrete use of on-screen human-like characters as persistent tutors or facilitators embedded into routine learning workflows rather than occasional content hosts.
Media And Content Production: ByteDance (字节跳动) and Zhejiang University (浙江大学) were described as promoting “long-video AI digital humans,” extending virtual on-screen figures beyond short hosting or brief broadcasts into longer narrative forms, knowledge expression formats, and entertainment arrangements that require more complex plots, characters, and multimodal generation processes. This development was framed as a production-side effort where digital humans are treated as durable screen characters suitable for extended-form content rather than limited-duration presenters.
Gaming And Community Mods: Bilibili (哔哩哔哩) was described as hosting demonstrations of “Mad Island” NPC mod content, with the mods characterized as community-distributed enhancements to non-player characters that alter in-game experience and social sharing around character behavior. This example treated game characters as modifiable synthetic entities whose perceived capabilities and roles are shaped by creator tooling and community circulation rather than only by the base game implementation.
September 16 News
Healthcare: Gushengtang (固生堂) applied medical specialty “AI avatars” as a capacity multiplier by launching the “Traditional Chinese Medicine AI Clone” program in June 2025 and scaling it to 10 disciplines by August, with explicitly cited coverage including oncology, dermatology, gastroenterology, and otolaryngology; the stated operational intent was to expand access to traditional Chinese medicine consultations, standardize expert guidance, and support international operations by cloning domain expertise into consistently available digital clinicians.
Tourism And Cultural Heritage: Aikesound (艾克声) deployed an AI Digital Human Guide System at CIFTIS that handled spoken queries, route planning, and exhibit narration as a venue-scale template for museums and cultural-heritage settings, while Huashi Shanbao (华视山宝) used rapid avatarization workflows, including one-minute digital-double creation and five-second photo pipelines, to generate hyper-real characters for cultural-tourism showcases inside large venues. Yaxin Culture Group (亚新文化集团) applied an MR “theatre-tour” system for guided, story-driven conference navigation, the Hubei Provincial Museum (湖北省博物馆) used the character “Chu Yin” as an artifact-anchored interactive host tied to museum iconography, and Mohuan Aoyou Technology (魔幻遨游科技) extended city-IP branding through “Aoyou Lion” as a persistent virtual figure for smart-city narrative promotion. China Unicom (中国联通) integrated digital-human interactive terminals into Wuxiang’s 5G+ smart-tourism build-out to harden virtual-guide experiences into scenic-area infrastructure, while China Mobile (中国移动) combined its Guangdong travel application “Yue You Yue AI” with its national “Xihe” anti-fraud assistant under a single virtual-agent umbrella spanning trip planning, safety prompts, and public education; China Telecom (中国电信) complemented this with an AI travel companion, Bangyan Technology (邦彦技术) contributed NuwaAI as an agentic digital human geared for on-site assistance, and Huayu Wenlv (华域文旅) deployed a Smart Touring Vehicle as a mobile, venue-scale embodiment for tourism operations.
Finance And Consumer Protection: Mashang Consumer Finance (马上消费金融) introduced a consumer-protection digital human positioned to broadcast rights education and streamline complaint outreach as a branded contact point inside a regulated service environment. China CITIC Bank (中信银行) deployed “Xiaoxin” as an intelligent wealth advisor aligned to the bank’s product journey, combining scripted guidance with interactive Q&A in retail finance.
Marketing And Retail Service Fronts: MCM integrated an actor-authorized face and voice into an AI shopping assistant in October 2024 as an identity-licensed approach to virtual retail staffing, and the same set of reports positioned this as a control and governance case where prompt discipline and persona drift become operational risks once a branded assistant is allowed to evolve beyond a narrow brief. Watsons (屈臣氏) was cited using “Qu Chenxi” as a virtual being for always-on promotion and service-front presence, Florasis (花西子) was cited using a brand-owned virtual human as a persistent spokesperson, and China Merchants Bank (招商银行) was cited using “AI Tips” as a comparable service-front virtual persona, collectively illustrating a spectrum from licensed likeness integration to fully brand-owned virtual being IP in customer-facing commerce contexts.
Enterprise Operations: CIIC Group (中国国际技术智力合作集团) presented an AI Digital Human Intelligent Recruitment Interview solution designed to standardize and personalize candidate screening at scale by generating role-specific virtual interviewers that balance consistency with adaptive questioning. FanTuo ShuChuang (泛图数创) targeted physical-space enterprise deployments by integrating virtual digital humans with smart wayfinding and dynamic-content engines for corporate showrooms, creating persistent brand-tuned presenters inside data-rich environments. Shanghai Kunzhiyi Artificial Intelligence Technology Co., Ltd. (上海坤智易人工智能科技有限公司) partnered with Zhangjiagang Leyu Town to run a 24/7 digital-human live-stream solution for e-commerce in which virtual anchors sustain product display, narration, and conversion beyond human shift limits. Tencent (腾讯) embedded a lightweight instantiation path for maintainable digital personae inside WeChat (微信) Official Accounts by enabling “Smart Reply,” where operators configure a “Digital Avatar” using a knowledge base and prompt so private-message replies remain consistent with the account’s voice without requiring a separate app build.
Media, Culture, And Public-Facing Narratives: Wondershare Technology (万兴科技) deployed the digital human “Xiao Lu” as a cultural-communication envoy for international outreach, while Henan’s AIGC digital-human reading companion was framed as a reading scaffold that promotes regional content (“Henan Stories”) through a virtual presenter aligned with curriculum and media production. In entertainment production, Creation of the Gods II was described as integrating thousands of shots combining performers with digital creatures and digital characters under senior VFX supervision to deliver dense action sequences, and talk-format programming was described experimenting with hybrid casting where Yu Qian and Cai Ming engaged the virtual persona “Nanako” as an on-screen participant.
Industrial And Utility Contexts: The “Wukong” power-operations digital human in Macau was framed as a persistent virtual operator used for safety briefings and live-line work education, extending digital humans beyond front-of-house roles into high-reliability operational environments where continuous guidance and compliance reinforcement are central to deployment value.
Legal And Governance Contexts: Beijing Internet Court (北京互联网法院) articulated a definitional frame in which virtual digital humans consist of an outward likeness plus a technical core, and clarified that portrait and voice rights are implicated when those elements are sourced from real individuals, setting a compliance perimeter for creators and operators of virtual beings. Shanghai’s policy measures were described as explicitly supporting digital humans alongside digital twins, VR, perception interaction, and spatial computing, positioning municipal-level enablement as an institutional on-ramp for the commercialization patterns reflected in the deployments above.
September 15 News
Healthcare: The “AI Health Manager AQ” is described as an AI doctor avatar used for preliminary diagnosis, medical report analysis, and skin testing, positioned as a personal health-management companion that combines medical expert knowledge with multimodal capabilities such as voice and visual analysis, while requiring engineering controls for data security and compliance with medical information processing rules.
Marketing and Commerce: Zhangjiagang’s AI Digital Live Broadcast Base deploys AI-generated digital humans for 24-hour livestream sales across a multi-account matrix, producing parallel content streams for categories including clothing, home appliances, and agricultural products to reduce staffing and venue costs; the same coverage frames conversion performance as dependent on interactivity and real-time Q&A rather than passive “display” avatars. Bangyan Technology (邦彦科技) is presented as offering a “zero-threshold” digital human creation platform for enterprises, providing a commercial character-production pipeline for livestreaming, talent-show contexts, and intelligent product explanations, including image definition, driving, and business-script assembly. In Jiangsu’s live e-commerce ecosystem, 24/7 AI digital-human hosts are described as operating at monthly costs of only a few thousand yuan, with the Jiangsu Live E-commerce Industry Innovation Service Base in Wuxi showing continuous virtual-presenter “bring goods” operations as a standardized pipeline; Baidu (百度) is described as promoting adoption via a reported 1-billion-yuan subsidy linked to a “new-generation persuasive digital human” aimed at SME onboarding. Tianyoude Liquor (天佑德青稞酒) is described as formalizing a named brand persona, “Li Ling’er,” and referencing a “Highland Barley Digital Human,” positioning named synthetic characters as durable brand assets integrated into an ongoing marketing stack rather than one-off campaign mascots.
Tourism and Culture: China Telecom (中国电信) is described as presenting an AI companion travel assistant and China Mobile (中国移动) Guangdong branch as deploying “Guangdong Travel Guangdong AI” as a digital-human front end for itinerary guidance and service Q&A in a travel-expo setting, while Bangyan Technology is also referenced in this domain through its NuwaAI agentic digital human positioned for hospitality-facing interactions and service guidance. Huayu Culture and Tourism (华域文旅) is described as using digital humans and MR guides on-site in expo and themed exhibition areas, combining virtual-image “floating” guidance with handle-based interaction to create gamified visit routes intended to improve viewing efficiency and immersion, and the same reporting links this to enterprise-side character-asset production for scenic-spot explanations and interactive demonstrations. Shantou is described as applying “Digital Human Customization” to performance contexts and “digital gong and drum interaction” to audience engagement, while Shandong exhibition programming is described as combining AI digital art, immersive performances, “Travel Photography Digital People,” and MR guiding to frame avatars as tour companions and docents. Guangxi Radio and Television Big Data Technology Co., Ltd. (广西广播电视大数据科技有限责任公司) is described as demonstrating “Cultural Digital People Dongdong,” combining knowledge graphs with LLMs to support cultural communication, and in Hong Kong, Cyberport (香港数码港) with the Hong Kong Design Institute (香港知专设计学院) are described as building a social-media-linked emotional digital human for live event interaction and sustained online engagement across venues and channels. A separate cultural reconstruction case attributes to the University of Hong Kong (香港大学) a virtual-human insertion and modeling approach that uses deep learning to extract the rhythm of Shen Congwen’s Xiangxi dialect and reproduce his tone and narrative style for historical restoration scenes, combining sound cloning, spoken-language synthesis, and scene synthesis for museums, school history museums, and documentary production while raising questions about authority and ethical boundaries in celebrity-voice digitization.
Government and Public Services: Gansu’s Cybersecurity Week is described as introducing “Lan Xiaoxin” as a co-hosting digital human used for public education and live events with distribution across China Gansu Net properties and major short-video platforms, illustrating institution-led scaling of consistent messaging through virtual presenters. Liaoning’s provincial medical-insurance authority is described as launching an AI digital human for public services and enrollment communication, integrating avatar-based assistance into citizen-facing workflows in higher-education settings and beyond, positioning a healthcare-adjacent administrative deployment rather than a clinical one. The same corpus also includes a governance case outside China in which Albania appoints an AI-powered digital human named “Diera” as Minister of Public Procurement as a pilot aimed at process transparency and public-service digitization, explicitly highlighting questions of responsibility boundaries, accountability, and explainability when a synthetic persona is placed in a formal public role.
Enterprise and Infrastructure Operations: In financial services, Industrial and Commercial Bank of China (中国工商银行) is described as using digital humans as “digital employees” for customer service, coordinated with payment, risk control, and wealth management process nodes to provide end-to-end support from consultation and guidance through to business processing. In the utilities domain, Guangzhou’s “Wukong” is described as a virtual operator that supports live-line power operations with round-the-clock availability, framing a synthetic persona as a persistent interface to operational knowledge and task support rather than a public-facing host. In XR-mediated collaboration, PICO (PICO/字节跳动系品牌) is described as offering a multi-user virtual space in which more than ten users can appear simultaneously as virtual avatars with voice, basic gestures, and laser-pointer instruction plus built-in interactive projects, positioned for a mixed set of immersive online meetings, teaching, and social entertainment where avatars function as identity presentation, guidance, and collaborative annotation tools.
Media and Education Deployments: The virtual host “Chu Yin” is described as appearing on stage at “Audio-Visual Yangtze River” Internet Communication Week as an AIGC-generated digital human used for on-site introductions and interactions, developed by the School of Design and the Key Laboratory of Light and Shadow Interactive Service Technology of Huazhong University of Science and Technology (华中科技大学设计学院与光影交互服务技术教育部重点实验室), with a technology chain covering image generation, speech synthesis, lip and expression driving, and integration with on-site broadcast-control workflows.
Memorial and Digital Afterlife: A subscription-based deceased-person avatar service is described as creating a conversational “digital avatar” by collecting personal voice and biographical materials to provide families with interactive memories after death, reporting more than 400 users and including cases where critically ill people recorded multiple conversations with family members during final life stages. The same description frames the system as integrating voice cloning and narrative modeling while emphasizing privacy authorization and consent requirements, including governance questions around pre-death consent and posthumous data handling.
September 14 News
Healthcare: HTC and its partners Hualien Tzu Chi Hospital and Tzu Chi University deploy “artificial intelligence virtual humans” for clinical communication, nursing training, and medical-education evaluation, framing these virtual beings as structured actors inside hospital workflows rather than promotional demonstrations. Baidu Smart Cloud (百度智能云) and Doshin Education use a multimodal large-model approach and the named “Xiling Digital Human” to implement “AI teachers” and smart tutoring functions that connect knowledge systems including search, encyclopedias, document libraries, and maps into teaching scenarios, while AQ uses an “intelligent body open platform” to create AI clones for more than 300 famous doctors and links nearly one million real doctors through the online consultation network of Good Doctor (好大夫在线) to support continuous 7×24 medical services; in parallel, regional “family doctor” deployments report more than 200 family-doctor AI clones to expand resident coverage and speed chronic-disease follow-up and routine consultations.
Government Services: The Chongqing Municipal Agriculture and Rural Affairs Committee embeds AI digital humans into a rural creator-training pipeline that explicitly ties content production to commerce outcomes, positioning digital humans as operational tools for rural branding and monetization rather than standalone mascots. Tianjin’s Beichen District Government Service Center runs the named consulting digital human “Xiaochen” to route citizens through procedures, Tianjin University of Technology extends an “AI digital human interaction platform” within its “Tianlige” campus IP to provide one-stop student-community services, and Beijing courts deploy and upgrade a human–computer-interaction AI digital human dedicated to juvenile protection services; in civic communication, Beijing’s CIFTIS presentations use the “Beijing Solution Ambassador” digital human for on-site explanation, Jiangbei District’s Cyberspace Administration in Chongqing uses the IP digital human “Shancheng Xia” for cybersecurity publicity, and Guizhou science-popularization programming uses a plush-styled digital human for public outreach.
Cultural Tourism And Museums: China Huashi Group (华视集团) showcases hyper-realistic digital humans with real-time multilingual interaction and AIGC video generation at a cultural tourism expo, paired with on-site digital-human creation experiences and “intelligent tour guide” deployments aimed at upgrading cultural pavilions and scenic-spot services. Guoshu Technology (国数科技) frames its coverage as spanning cultural pavilions, ethnic tourism, and virtual fashion shows while extending into metaverse digital-human content production and performances, and multiple public-facing cultural cases present digital humans as standard explanatory interfaces, including the “Beijing Plan Promotion Officer” digital human used for exhibition explanation, Changsha cultural venues using AI digital humans to narrate Huxiang history, and the Henan Pavilion at CIFTIS assigning a digital human to explain publishing-driven culture export; related “AI + culture–tourism” implementations include Shantou’s “Tide-Rising Yingge: AI Digital Space Experience Hall” and Guangxi’s “Products Going Overseas” showcase using AI digital humans to present regional goods and intangible-heritage items.
Education: Zhejiang undergraduate arts education introduces embodied digital-human versions of Li Bai and Du Fu for first-person poetic commentary, using avatarized historical figures as instructional presenters rather than passive visualizations, and Hubei’s public-facing “expert” digital humans similarly present social-science material in an accessible, persona-led format. In higher-education services, Tianjin University of Technology’s “Tianlige” campus IP incorporates an AI digital human interaction platform for student-community support, while Baidu Smart Cloud (百度智能云) and Doshin Education’s “Xiling Digital Human” positions AI teacher embodiments as tutoring agents connected to broader knowledge systems; at the skills-training layer, Ningxia’s Yinchuan competitions include building “digital human live rooms,” operationalizing avatar hosts as teachable production competencies within integrated marketing and content workflows.
Marketing And E-Commerce: Keyiyun (科易云) markets “Digital Human Live Broadcast” as always-on avatar anchors designed to remove human staffing bottlenecks in e-commerce operations, with core functions centered on persistent hosting and scalable livestream continuity rather than one-off campaigns. Leapmotor (零跑汽车) uses the “Tiger Chief Digital Human” as a hyper-realistic live-broadcast spokesperson for vehicle marketing, indicating a brand-owned avatar deployed across launch communication and sales-enablement content, and the broader tooling landscape described as “Digital Human AI livestreaming” emphasizes avatar generation, real-time interaction, and content generation to support brand livestreaming and e-commerce promotion while reducing marginal production costs and enabling event scaling; the Chongqing rural creator-training pipeline also ties AI digital humans directly to commerce conversion for agricultural branding, demonstrating a second, non-automotive pathway where digital humans function as growth infrastructure for local products.
Enterprise And Financial Services: Bank of China (中国银行), through its Guangxi branch, introduces “Zhongxiaogui” as a multilingual voice-interactive digital human oriented to SME customers, using the digital human as a front-end service interface for triaging requests and guiding operations across languages in a concrete banking setting. In adjacent enterprise deployment narratives, Mingtou Technology (明拓科技) is identified as Sichuan’s Intelligent Digital Human Engineering Technology Research Center and is positioned as brokering university–enterprise collaboration to translate digital-human research into applications, while Shanghai’s Bund Conference reporting frames digital-human and avatar interfaces as HR-facing front ends for recruitment and candidate job search, indicating continued use of embodied agents as interaction layers in enterprise service processes.
Media And Entertainment: Chengdu’s live-action sci-fi wuxia film “Emei” explicitly integrates AI digital humans within a 2045 Emei-Jinding narrative, treating digital humans as in-world performers or narrative agents rather than post-production novelty. Nanning Fengzhi Culture Communication Co., Ltd. (南宁凤之文化传播有限公司) presents the AIGC short-drama creation tool “Ju Xiaobai” as a production pipeline for scripted content generation, and Guangxi Broadcasting Big Data Technology Co., Ltd. (广西广电大数据科技有限公司) demonstrates “Cultural Digital Human Dongdong,” described as knowledge-graph- and large-model-based, for culturally grounded narration that formalizes a named synthetic character as a presenter aligned to regional cultural storytelling.
Governance And Research Evidence: WeChat (微信) launches a “digital avatar” capability for official accounts through “Smart Reply,” positioning avatarized brand identities as always-on responders that match content and handle uninterrupted interaction rather than relying on ad hoc human operators. Governance debates noted alongside these deployments include public discussion of legal identification rules for “virtual digital people” and questions about platform responsibility boundaries, identity authentication, and content compliance when virtual people participate in social activities, transactions, and performances, while empirical research on virtual influencers and AI digital human anchors in live commerce links concrete design variables—parasocial relationship strength, influencer-product fit, expectation confirmation, and responsiveness—to purchase intention via trust, and separate film-focused work on “AI actors” reports that labeling performers as AI changes audience evaluation and watching intentions, establishing disclosure and reception effects as operational considerations for synthetic-character casting and presentation.
September 13 News
Education: Moore Threads (摩尔线程) is positioned in classroom deployment through Wuxi’s AIBOOK-based training base, where its hardware supports a Spark-blackboard-integrated AI teaching digital human used for primary-to-high-school courses, framing a virtual teacher as a standardized front end for lessons and demonstrations across grade levels. Shandong Digital Human Technology Co., Ltd. (山东数字人科技有限公司) is supplying a digital anatomy and morphology experimental teaching platform to Gansu Medical College, linking digital-human simulation to hands-on laboratory instruction where anatomy and morphology content can be explored through a controlled virtual interface rather than only through physical materials.
Enterprise And Finance: China National Petroleum (中国石油天然气集团有限公司) uses a digital human to popularize its Kunlun large model, pairing a visible explainer persona with corporate AI capabilities in a context that resembles internal evangelism and external brand communication for an enterprise foundation model. China Construction Bank (中国建设银行) demonstrated virtual digital humans built from real-employee models using motion capture at CIFTIS, and combined these agents with one-stop devices supporting updated housing-fund co-branded cards, presenting a regulated, transaction-adjacent service flow in which a bank-branded digital human is integrated into a booth-style customer journey. Ant Group (蚂蚁集团) is linked through Alipay’s launch of the Xiaoye employment assistant, which applies a large model with a digital-human front end to job services, positioning an avatar interface as the visible guide for employment-related tasks inside a high-frequency consumer platform.
Platforms And Tooling: ByteDance (字节跳动) anchors a creator stack that includes Seedream 4.0 for high-fidelity image generation and CapCut’s AI avatar generator for turnkey character creation, lowering the production threshold for creating avatars that can be reused across social venues and content formats. Tencent (腾讯) is represented through WeChat Official Accounts tooling that enables operators to configure a “digital twin” capable of answering messages around the clock with a tailored knowledge base and style, describing a platform-level pathway for persistent branded personas used in support, marketing, and community engagement. Diandong Technology (电动科技) is described as establishing a joint laboratory with HKUST (Guangzhou) focused on intelligent-agent systems and high-fidelity 3D digital-human pipelines, emphasizing an end-to-end technical pathway from controllable generation and reconstruction research to commercialization-oriented digital-human production.
Media And Entertainment: Shanghai United Media Group (上海报业集团) is described as operating an AI super-platform and a “digital human matrix” across outlets including The Paper, Jiemian, and CLS, indicating an organizational approach where digital humans are treated as a repeatable production asset across a converged media portfolio. Chengdu Film City (成都影视城) is linked to the signing of the live-action sci-fi wuxia drama “Emei,” where AI digital humans are integrated into serial production, framing synthetic performers and digital characters as part of the production pipeline rather than isolated demonstrations.
Culture And Tourism: Qingdao Metro Group (青岛地铁集团) is associated with an award-winning brand story that integrates digital-human elements, using a virtual character layer to support corporate image-building for a public-transport operator. Sichuan Mingtou Technology Co., Ltd. (四川明投科技有限公司) and SAT Entertainment Group Sdn. Bhd. are tied to the “Citywalk Intelligent-Agent Digital Human Development Project,” which targets urban exploration companions designed to support wayfinding, local discovery, and destination storytelling as users move through city-scale experiences. The Summer Palace AI digital human is framed as a smart customer-service agent for cultural-heritage engagement at a major national site, and CIFTIS coverage also places “AI digital guides” within exhibition contexts in Beijing, treating virtual humans as consistent, personable interfaces for narration, routing, and visitor assistance in cultural venues.
E-Government: Tianjin’s Beichen District deploys Xiao Chen as a 24/7 administrative assistant designed for instant responses in citizen services, presenting a digital human as a front-desk surrogate intended to reduce queueing and standardize answers for routine inquiries. The Rizhao Economic and Technological Development Zone Procuratorate’s youth program uses a virtual instructor for rule-of-law education, applying an avatar interface to scripted legal-literacy delivery where predictable question sets and structured guidance are prioritized over open-ended interaction.
Healthcare: Hangzhou’s traditional Chinese medicine technology demonstrations include digital humans alongside a smart TCM pharmacy as part of diagnostic and therapeutic workflows, suggesting an avatar layer used for pre-visit guidance, intake support, and patient education within integrative medicine settings. A separate research example describes a VR-delivered avatar developed at the Technical University of Munich to communicate with patients during autonomous ultrasound, reporting stress mitigation through an interactive, humanized interface that explains procedures and maintains a bedside communication channel during automated diagnostics.
Telecommunications: China Unicom (中国联合网络通信集团有限公司) deploys the AI holographic digital human Tongtong built on its Yuanjing large model and incorporating DeepSeek (深度求索), presenting a carrier-scale stack where a named digital human is used in public exhibitions and service contexts as the visible, interactive interface to underlying model capabilities. The deployment framing emphasizes a branded, demonstrable virtual presence that can be reused across events and service encounters while remaining anchored to a telco-controlled foundation model and an integrated external model component.
September 12 News
Healthcare: GuShengTang (固生堂) is described as using “AI clones” of eminent Traditional Chinese Medicine experts by modeling ten physician personas spanning eight core TCM domains, positioning these digital human replicas as a way to extend specialist access where expertise is unevenly distributed and linking the deployment to reported mid-2025 profit growth. China Telecom (中国电信), via its Chongqing Telecom branch, is described as applying the “Xingchen Big Model” to medical-insurance services by launching a “Medical Insurance Think Tank,” a “Medical Insurance Digital Human,” and a “Deep Thinking Smart Doctor,” with digital humans operating at the front end for consultation and acceptance while linking to medical auxiliary decision-making and process management as a knowledge-center entry point for medical-insurance workflows; Chuanshen Language Network is described as exhibiting interactive digital humans alongside the Rendu big model and the Suwen Chinese Medicine big model to support Chinese-medicine knowledge Q&A and multilingual dialogue, with the same interactive digital-human setup also used for adjacent public-service and visitor-interaction scenarios.
Education: Nanning Normal University (南宁师范大学) is described as using “AI digital human teaching assistants” to lead students in intangible cultural heritage practice, with the digital humans responsible for task decomposition, skill demonstration, and process accompaniment as a bridging teaching-assistant node between classroom instruction and hands-on practice. Yuanfudao (猿辅导) is described as extending digital-human deployment into family-education workflows through a virtual-image AI teaching assistant that acts as a continuous companion across pre-class, in-class, and post-class phases, emphasizing continuity of guidance rather than a single-session tool. Bairong Cloud Innovation (百融云创) is described as reusing a unified digital-human setting across domains by trialing “digital human all-in-one machines” on campuses in Shenzhen and Shijiazhuang that use a “Little Monkey King” character to provide students with emotional guidance and suggestions, while the same underlying digital-human platform is also positioned for enterprise risk-control and marketing applications.
Cultural: China Unicom (中国联通), via its Heilongjiang Unicom branch, is described as using an AI assistant with a virtual digital human image inside its “One-Code Travel” smart-tourism offering to deliver scenic-area explanations, route guidance, and consultations through smart screens and mini-programs, framing the digital human as the interaction layer that completes a closed loop for on-site tours. Chuanshen Language Network is also described as using “interactive digital humans” with VR at the China International Fair for Trade in Services to let visitors “travel online” to Hubei landmarks and interact with the named digital human “Chu Shiqi,” positioning the character as a regional image and cultural-communication interface. Guangxi Radio and Television Big Data Technology Co., Ltd. (广西广电大数据科技有限公司) is described as deploying the “Cultural Digital Human Dongdong,” which integrates knowledge graphs to deliver cultural-theme Q&A and explanations as a dialogue front end for cultural communication, while multiple exhibitions including CIFTIS and the World Manufacturing Conference are described as using digital human tour guides and an “AI Digital Human Guide + Central Axis Check-in” format, including attention-grabbing staging such as “digital humans suspended in the air” to increase visibility and interactivity in visitor-facing cultural exhibition spaces.
Marketing: Leapmotor (零跑汽车) is described as using its chief digital human “Tiger” to conduct hyper-realistic live broadcasts that handle new product introductions, fan interaction, and brand-personalized narratives, positioning the digital human as a scalable spokesperson for long-tail automotive marketing content. Chengdu Mingtu Technology Co., Ltd. (成都明图科技有限公司) is described as using cartoon digital humans in offline exhibition halls for real-time audience greeting, brand welcome, and crowd guidance, emphasizing on-site interaction rather than purely online distribution. Nanning Gaofeng Culture Communication Co., Ltd. (南宁高峰文化传播有限公司) is described as launching the AIGC short-play creation application “Ju Xiaobai,” which generates scripts, shots, and characters to assist short-play production, while Hangzhou Shangcheng’s “Digital Human Anchor Application” is described as providing virtual appearance and broadcasting capabilities for media communication so digital humans can replace or supplement human anchors for city image promotion and event information release, extending digital-human hosting into municipal-facing media workflows.
Enterprise: Industrial and Commercial Bank of China (ICBC) (中国工商银行) is described as fielding a “Housing Provident Fund digital human” within a one-stop service booth at CIFTIS to streamline citizen interactions, while ICBC Credit Suisse Fund (工银瑞信基金) is described as staging an anti-money-laundering exhibit aimed at public compliance education, together illustrating customer-journey and public-education roles for institutional digital humans in finance. Shanghai Shiwei Technology Co., Ltd. is described as promoting a metaverse digital human system and a metaverse meeting system for corporate functions spanning design, training, marketing, and operations, presenting the digital-human layer as a multi-department cost-reduction tool rather than a single-purpose interface. In operational and infrastructure contexts, China Mobile (中国移动), via Shanghai Mobile, is described as running a joint innovation with Nokia Bell that introduces digital humans into network-security sites for promotion and communication, with the same platform’s knowledge and interaction capabilities described as reusable for teaching and training scenarios; Guangzhou Digital Technology Group is described as demonstrating a smart-airport solution in airport scenarios, and Yuntong Intelligence is described as embedding “flight display + digital humans” into flight-display systems to provide broadcasts and inquiries, positioning digital humans as an information and service interface in transportation operations.
Ethical Contexts: The Beijing Internet Court is described as recognizing original virtual digital human images as protectable art and holding that unauthorized creation of a natural person’s AI virtual likeness infringes personality rights, alongside establishing protection for natural voices in an AI voice-rights case, together framing a rights-and-liability baseline for digital human production involving likeness and voice. A concrete enforcement example is described in which a Hangzhou company was ordered to pay 20,000 yuan after deploying an AI-cloned voice of a former employee in a virtual role and publishing related content on Bilibili (哔哩哔哩), tying virtual-character deployment choices directly to workplace-related consent, voice cloning, and platform-facing publication risk.
September 11 News
Enterprise: Future Secure AI, backed and piloted by Macquarie, is placing avatar workers into finance and HR workflows as agentic labor, framing these virtual beings as operational “workers” embedded in routine corporate processes rather than as novelty interfaces. analogai.net promotes “Silvio” as a hireable digital human HR interviewer positioned to run real screening flows, while Mirako presents the operational digital human “Maya” as an upgrade path from a static model to an agentic actor capable of interactive, goal-directed exchanges. Alibaba DingTalk (钉钉) is described as the enterprise layer through which Shunfu Precision is piloting “digital human marketing” for acquisition and service, and PhaseCore Technology is positioning an in-car digital human as a conversational interface for vehicle interaction, indicating a convergence between customer engagement, sales enablement, and persistent front-office personas. QMMM Holdings Ltd. is extending from virtual-avatar media into a blockchain-AI data marketplace and avatar-driven virtual apparel, while Marina Bay Sands uses a metaverse environment in which conference delegates appear and navigate as avatars with in-world chat coordination, showing that avatar-mediated participation is being deployed at venue scale for events and internal coordination.
Education: Boise State University integrates AI avatars into course delivery in TEAM303 as instructional agents aligned with broader classroom diffusion of generative AI tools, demonstrating a concrete deployment context where an avatar is part of instructional delivery rather than a separate novelty layer. Nanning Normal University launched a “Labor Education AI Digital Human Teaching Assistant System” for intangible cultural heritage courses after three months of development and live audio demonstrations, and NetDragon is described as showcasing an “Interactive AI Digital Human” at the United Nations 2025 Digital Learning Week with a hardware stack that includes Promethean ActivPanel 10 and ROKID AR equipment for multimodal classroom and training interaction. Gansu Province is described as refining a “Cultural and Museum Digital Human” commentary series with eight released episodes for exhibitions related to Fuxi culture, while teacher Hu An is described as applying “Digital Human Narrative Technology” to train virtual avatars of “Digital Elders” to preserve rural memories, extending educational deployment into heritage transmission and community memory practices.
Marketing: Vodafone Germany is running an AI influencer on TikTok for campaign work, and this is explicitly linked to the broader AI-influencer category through reference to Lil Miquela as a recognizable benchmark, keeping the focus on influencer-style synthetic characters as marketing channels. Baidu (百度) is described as disclosing that digital human live-stream conversion rates exceed those of real people and pairing this with million-dollar incentives, integrating its Wenxin Big Model X1.1 Deep Thinking Model into ultra-realistic live-broadcast performance and content generation as a commercialization stack for digital human live commerce. Keyiyun is presented as supplying a “Digital Human Live Broadcast All-in-One Machine” and a “Digital Human Live Broadcast System” for integrated broadcast control and immersive commercial solutions, while Wenxin Exhibition Service Co., Ltd. is described as introducing “Digital Human Advertising” and an “Immersive Experience Space” at the Shanghai International Printing and LED Digital Signage Joint Exhibition. The Guangdong Live E-Commerce Association is described as incorporating “AI Digital Human” into live e-commerce training and ecosystem collaboration with platforms, parks, and governments, tying talent development and operational playbooks to rollout.
Entertainment: ABBA Voyage continues to run concerts built around digital avatars of the band at a purpose-built Stratford venue with ticketing extending into 2026, showing an avatar-led performance as a sustained commercial format rather than a one-off experiment. The short play If Qin Bamboo Slips Could Talk is presented as using the digital human “Xi” to interact with children across time and space on stage, combining historical interpretation with character performance and audience interaction in a youth-education adjacency that still functions as a theatrical case study. Suno AI and Udio are described as powering multilingual virtual singers by generating melodies and harmonies, while HeyGen and Runway are described as enabling brand-ready virtual hosts and animated explainers that push synthetic character production into repeatable pipelines for commercial media output.
Gaming: Thirdverse and Sanrio’s Hello Kitty Skyland is described as moving social-VR fandom into closed beta with avatar-based play, positioning the avatar layer as the core of social participation and identity. Zeta’s AI characters are described as driving teen adoption in Korea through intimacy-scripted interactions, while Tavern AI is described as providing an open-source path to craft personalized AI sidekicks, and the Nano Banana AI model is described as supporting narrative-driven companion experiences, collectively framing synthetic characters as interactive, relationship-oriented entities inside entertainment-adjacent ecosystems. Reddit discussions around Elite Dangerous, Critical Role, and Hitman are described as emphasizing NPC personality, consistency, and believability, and Bilibili (哔哩哔哩) is described as hosting a “Mad Island” NPC mod showcase that demonstrates user-driven pipelines to alter NPC behavior and presentation through a distribution-and-feedback loop.
Healthcare: Replika is described as exemplifying persistent, personalized AI companionship optimized for relationship continuity and emotional scaffolding rather than task throughput, making the virtual being’s design stance and memory-like continuity part of the core function. Born is described as raising US$15 million to advance “social” AI companions including the virtual pet Pengu and to expand its AI-character engine and youth-focused social product, framing companions as supports for loneliness reduction through a consumer social-product pathway. The Technische Universität München ultrasound avatar is described as a clinical interface that can increase patient confidence in imaging procedures through consistent presentation and intelligible guidance, narrowing the healthcare case to trust, adherence, and bedside-manner-like interaction.
Culture: Harsha P. Deka is described as offering fully rigged custom 3D avatars for VTubers and virtual-world users with optional physical figurines, linking avatar identity to merchandise-ready outputs as a practical toolchain and commerce extension. Google is described as enabling viral 3D digital avatars of public figures through consumer-facing features, with meme-driven appearances of public-official avatars illustrating rapid spread when creation friction is low, and Nike’s CloneX series is described as anchoring the linkage between brand activity, avatar wearables, and metaverse-oriented digital collectibles. Codansha’s project with Tohoku University is described as co-developing the AI character RD with fans through a training-through-fandom model where community interaction continuously shapes the character’s knowledge and behavior, presenting a participatory IP-to-virtual-being conversion pipeline.
Enterprise Operations in China: Jiangxi Unicom and Luxshare Intelligent Manufacturing are described as using “digital human instructors” in factories to standardize skills training and improve operational efficiency, and Guangzhou Industrial Control Group is described as promoting collaboration with universities and research institutions to cultivate digital-human-related talent as an industrialization pathway. Hubei’s traditional clothing industry is described as introducing “AI digital human anchors” alongside smart hanging and automatic sorting systems to connect production and marketing, framing the digital human as a front-end sales and promotion function embedded in manufacturing modernization rather than as a standalone media asset.
Technology and Platforms: Meta is described as preparing smart-glasses features that let users customize personalities, voices, and behaviors for AI-driven avatars, linking avatar personalization to wearable deployment contexts and everyday device ecosystems. Copresence is described as converting phone scans into lifelike 3D avatars and contrasted with Meta’s studio-intensive Codec Avatars, while Holoworld AI (HOLO) is described as enabling creation of intelligent, task-performing virtual-being agents across platforms and distributing via Binance airdrops, placing virtual beings into a blockchain-linked distribution and incentive context. ElevenLabs is described as supplying voice AI used across audiobooks, dubbing, gaming, and avatars, positioning voice synthesis as a cross-domain enabling layer for embodied or character-like agents rather than as a standalone voicebot. MetaHuman templates for Unreal Engine provide rapid, style-consistent character creation inside the engine ecosystem, and a MetaHumanForMaya note indicates that Maya 2025.3 integration may require explicit compatibility updates; in parallel, the VHToolkit (VHTK) at ict.usc.edu is described as a mature legacy resource for building interactive virtual humans, indicating continuity between earlier virtual-human frameworks and current mass-market creator tooling.
Legal and Ethical Contexts: Senate testimony is described as flagging the need for safeguards where avatar abuse in VR can cause real trauma to children, treating avatar-mediated harm as a concrete risk scenario with child-safety implications rather than an abstract concern. Beijing decisions are described as recognizing personality-rights coverage for AI-generated virtual images and voices, and a Hangzhou case involving the virtual digital human “Ada” is described as underscoring that motion-capture performances are protectable and that TikTok can function as a dissemination vector in unauthorized use, establishing early patterns for enforcement and platform responsibility. The Hangzhou Court and the Hangzhou Intellectual Property Court are described as sorting adjudication ideas and behavioral boundaries for new disputes involving digital human images, content generation, and commercial use, including questions framed around one-click generation and infringement, pointing to emerging compliance frameworks grounded in case handling rather than policy aspiration.
September 10 News
September 09 News
September 08 News
September 07 News
September 06 News
September 05 News
September 04 News
September 03 News
September 02 News
September 01 News