Beijing’s Action Plan for Promoting Innovation and Development of the Digital Human Industry (2022–2025), 北京市促进数字人产业创新发展行动计划（2022-2025年), frames “digital humans” as a core Internet 3.0 opportunity and sets out a city policy to build a competitive technology system, active innovation and business models, and a “tolerant and prudent” governance approach that balances development with security and rules with innovation. It positions the work within national digital-economy strategy and argues for using Beijing’s international science-and-technology innovation strengths to create an industry innovation hub that supports the city’s broader digital-economy ambitions.
The plan’s 2025 targets focus on industrial scale, leading firms, shared platforms, and demonstrable deployments. It calls for an industry size exceeding RMB 50 billion, the cultivation of 1–2 leading firms with annual revenue above RMB 5 billion and 10 firms above RMB 1 billion, breakthroughs in key core technologies, and the establishment of 10 university–enterprise joint laboratories and enterprise innovation centers. It also targets building at least five shared-technology platforms in areas such as cloud rendering, interaction driving, intelligent computing, data openness, and digital-asset circulation, creating 20 benchmark application projects in sectors including culture and tourism, finance, and government services, and building at least two specialized industrial parks and bases.
On the technology roadmap, the plan prioritizes end-to-end capability across the digital human production chain, from foundational software to core algorithms and enabling hardware. It supports enterprises, research institutes, and key laboratories to pursue technical breakthroughs in modeling software, rendering engines, and physics engines, and highlights algorithms such as ray tracing, deep-learning super sampling, facial modeling, posture and expression simulation, and high-fidelity video stream compression. In parallel, it calls for advances in foundational hardware such as compute chips, sensors, optical components, and batteries to strengthen an indigenous and controllable technology chain.
The plan also emphasizes improving production and operational tooling so digital humans can be made and deployed more efficiently and safely. It supports small-batch validation of new technologies and products with subsidies where projects meet “new infrastructure” promotion conditions, and it encourages development and deployment of VR/AR/MR terminals and display solutions such as glasses-free 3D and holographic imaging. It explicitly promotes the integration of AI functions including speech recognition, speech synthesis, and natural language understanding, and it calls for “smart brain” and affective computing algorithms to improve interaction, alongside low-code editing and operations platforms and a unified security tooling and testing/evaluation system for digital human software and hardware.
For shared infrastructure, the plan backs high-precision, low-latency cloud rendering platforms and edge computing tied to 5G networks, and it supports data-capture facilities such as motion-capture systems and XR studios. It also proposes building a DEM (digital elevation model) platform and exploring the opening and sharing of spatial data for real streets and buildings to provide a “space data base” for digital human interaction and presentation, with eligible projects able to receive support under “new infrastructure” policies. In addition, it advances a standardization program that encourages firms and social organizations to lead international and Zhongguancun standards, provides rewards for new international standard proposals and newly issued Zhongguancun standards, and supports the development of standards at international, national, industry, local, and group levels, including solution promotion aligned with self-developed standards.
Application policy is organized around service-oriented deployments, performance-oriented innovation, and a data-factor market for digital human assets. For service-oriented digital humans, it supports experiential “test-and-validate” scenarios, promotes consumer-facing uses such as e-commerce live streaming and “streaming media” production with full-scenario XR venues, and specifies finance use cases including remote visual teller services, video-based review and signing, digital-bank intelligent customer service, and in-branch “digital bank clerk” roles, while also promoting deployments at landmarks, museums, theaters, stadiums, and patriotic education sites and encouraging preferential traffic/data packages for distinctive applications. For performance-oriented digital humans, it aims to connect technical and artistic talent to incubate IP and a creator economy, expand interactive formats across film, television, online video and short video, explore fast-track approvals for digital-human variety shows, concerts, livestreams, and films, and support advertising endorsements and brand building with stronger operational management aligned to “positive energy” communication goals, including offline experience scenes in communities, malls, plazas, and commercial streets. For the data-factor market, it proposes integrating digital humans into a national cultural data network and service platform, enabling compliant data trading with stronger regulation, exploring blockchain-enabled marketplaces for models, skins, textures and related data elements, and piloting digital-asset valuation via data zones and a big-data exchange.
To strengthen the ecosystem, the plan calls for improved intellectual-property and copyright mechanisms, including cross-channel/platform/device traceability, rights protection, and services such as registration and certification covering digital humans and related works that may include software, real-person likenesses, and real-person voices, as well as protections for design patents and trademarks. It also sets out an industrial clustering approach based on parks and bases that provide support policies and incubation services, and it assigns district-level development directions, including Tongzhou for infrastructure clustering, Daxing for premium consumer applications leveraging new media base resources, Shijingshan (Shougang Park) for exhibition and experience scenes, Chaoyang for art creation and industrial incubation to form a municipal digital human base, and Fengtai for creator-economy development connected to modern finance and business formats. It further promotes a tiered enterprise pipeline focused on “specialized and sophisticated” firms and “little giant” cultivation along the industry chain, alongside talent development through industry–education integration pilots and complementary measures for leading talent, and it outlines financing support via government-guided funds, the Beijing Equity Exchange “Specialized and Sophisticated Board,” and IPO service strengthening.
In governance and safeguards, the plan proposes building collaborative innovation consortia across government, industry, academia, research and application under unified IP management frameworks, using open scenario projects to concentrate innovation, and creating international exchange platforms through major Beijing-hosted conferences and forums to promote Chinese technologies and standards and encourage “going global” approaches. It also requires multi-layer risk prevention focused on data security and personal information protection, content safety management and service-provider responsibility, and enforcement against illegal activity such as financial fraud and illegal fundraising conducted using digital humans. Finally, it calls for exploring a humanistic-ethics arbitration mechanism with industry association participation and for developing a diversified regulatory pathway combining law, markets, code architecture, and social norms.