China Digital Virtual Person Information Services Measures (Draft)

Overview

The Administrative Measures for Digital Virtual Human Information Services (Exposure Draft), hereafter referred to as the "Draft Measures," were issued by the Cyberspace Administration of China (CAC) on April 3, 2026. This significant regulatory initiative aims to foster the healthy development and standardized application of digital virtual human information services within China, while simultaneously safeguarding the legitimate rights and interests of citizens, legal persons, and other organizations. The Draft Measures represent a proactive step by Chinese authorities to address the emerging challenges and opportunities presented by rapidly advancing artificial intelligence technologies, particularly in the realm of AI-generated digital humans. The regulation seeks to strike a balance between encouraging technological innovation and ensuring robust oversight to mitigate potential risks associated with the misuse of these sophisticated digital entities.

The scope of the Draft Measures extends to all internet information services provided to the public within the territory of the People's Republic of China through digital virtual humans. This comprehensive approach underscores China's commitment to establishing a holistic governance framework for AI, building upon existing regulations concerning network security, data protection, and personal information. The introduction of these Draft Measures signifies a crucial phase in the evolution of virtual digital images from mere technological exploration to an era of stringent compliance regulation, reflecting a broader national strategy to ensure that AI development aligns with societal values and national security interests. The public comment period for these Draft Measures concludes on May 6, 2026, allowing stakeholders to provide feedback before finalization.

Definitions

The Draft Measures provide a clear and comprehensive definition of a "digital virtual human." According to the regulation, a digital virtual human is characterized as a virtual digital image that exists within a non-physical environment. These entities are created and animated using advanced technologies such as computer graphics, digital image processing, or artificial intelligence. They can be driven either by real persons (e.g., motion capture, voice acting) or by computational algorithms, and are designed to simulate human appearance. Crucially, digital virtual humans are endowed with characteristics such as voice, behavior, interactive capabilities, and even distinct personality traits, making them highly realistic and capable of complex interactions.

This definition is critical for delineating the scope of the regulation, ensuring that all forms of AI-generated or AI-driven human-like virtual representations are covered. The inclusion of elements like "interactive capabilities" and "personality traits" highlights the regulatory focus on the sophisticated nature of modern digital humans, which can engage with users in nuanced and potentially influential ways. The Draft Measures differentiate between digital humans created based on real individuals and those generated entirely by algorithms, acknowledging the diverse origins and applications of this technology. This foundational definition underpins the various requirements and prohibitions outlined throughout the regulation, particularly concerning personal information protection, content moderation, and ethical use.

Governance and Institutional Framework

The governance and institutional framework for overseeing digital virtual human services in China is designed to be multi-layered and collaborative, with the Cyberspace Administration of China (CAC) at its core. The CAC is designated as the primary regulatory authority responsible for coordinating the overall governance of digital virtual human services. This central role reflects the CAC's broader mandate in managing internet information content and cybersecurity across the nation. The Draft Measures empower the CAC to lead the formulation, implementation, and enforcement of policies related to digital virtual humans, ensuring a consistent national approach to this rapidly evolving technological domain.

Beyond the CAC, a range of other state departments are tasked with supporting the supervision and management of digital virtual human services, each according to their specific areas of expertise and jurisdiction. These include departments responsible for telecommunications, public security, healthcare, market regulation, media, film, and copyright. This collaborative model aims to ensure comprehensive oversight, addressing the diverse legal, ethical, and technical implications of digital virtual humans across various sectors. Furthermore, local internet authorities are mandated to enforce the same standards at a regional level, ensuring that the regulatory framework is applied consistently throughout China and adapting to local contexts where necessary. This distributed but coordinated governance structure is intended to provide robust and effective oversight of digital virtual human services.

Key Focus Areas

The Draft Measures delineate several key focus areas to ensure the healthy and responsible development of digital virtual human services. A paramount concern is the protection of minors, with strict prohibitions against services that could induce addiction or excessive consumption among children. Specifically, digital humans are forbidden from offering virtual relatives, romantic partners, or emotionally intimate relationships to users under 18 years of age. The regulation also bans content that promotes unsafe behavior, extreme emotions, moral violations, or harmful habits, and prohibits any digital virtual person service that could negatively affect a child's physical or mental health. This child-first approach reflects China's broader commitment to safeguarding the well-being of its younger population in the digital age.

Another critical focus is the protection of personal information and fundamental rights. The Draft Measures mandate explicit consent before using an individual's likeness, voice, or other personal data to create a digital human. This requirement extends to sensitive personal information, which necessitates separate consent and prominent notification of processing purposes and potential impacts. The regulation also strictly prohibits activities that infringe upon rights such as reputation, privacy, and intellectual property. Furthermore, the Draft Measures aim to prevent the misuse of digital humans for illicit activities, such as bypassing identity authentication systems (including facial or voice recognition), illegally registering or trading online accounts, or engaging in fraud and false promotion. Content restrictions are also significant, with prohibitions against disseminating information that harms national security, promotes extremism or violence, incites discrimination, distorts historical figures, or contains sexual innuendo, cruelty, or horror.

Implementation Framework

The implementation framework for the Administrative Measures for Digital Virtual Human Information Services is structured to establish a comprehensive compliance ecosystem, integrating various regulatory mechanisms and stakeholder responsibilities. A fundamental aspect of this framework is the emphasis on transparency and clear identification. Service providers, users, and content distribution platforms are explicitly required to display prominent labels, such as the characters "数字人" (digital human), throughout the entire duration of a digital virtual human's display. This labeling must comply with existing national regulations concerning AI-generated synthetic content, ensuring that users can easily distinguish between human and AI-generated interactions. This measure is crucial for preventing deception and maintaining public trust in digital services.

The framework also assigns clear responsibilities to different entities involved in the digital virtual human ecosystem. Service providers are expected to establish robust internal management systems, content review mechanisms, and technical prevention measures. This includes strict protocols for the collection, use, and storage of data, ensuring legal compliance and preventing data leakage or misuse. For digital virtual human services that possess public opinion attributes or social mobilization capabilities, the Draft Measures stipulate that providers must undergo algorithm filing and security assessment procedures, aligning with the requirements of the Internet Information Service Algorithm Recommendation Management Provisions. This tiered approach to regulation ensures that services with higher potential societal impact are subjected to more rigorous scrutiny and oversight, fostering a secure and responsible digital environment.

Monitoring and Evaluation

Monitoring and evaluation are integral components of the Draft Measures, designed to ensure ongoing compliance, assess the effectiveness of the regulations, and adapt to the dynamic landscape of AI technology. The Cyberspace Administration of China (CAC) and relevant competent authorities are empowered to conduct supervision and inspection activities concerning digital virtual human services within their respective jurisdictions. This oversight includes regular checks on service providers, technology supporters, service users, and platforms that disseminate network information content, ensuring adherence to the stipulated requirements regarding content, data handling, and user rights. The multi-departmental approach to governance, involving telecommunications, public security, market regulation, and other agencies, facilitates a comprehensive monitoring strategy that covers various aspects of digital virtual human operations.

Furthermore, the Draft Measures implicitly encourage a continuous evaluation process, particularly through mechanisms like algorithm备案 (algorithm filing) and安全评估 (security assessment) for services with significant public impact. These processes serve not only as pre-emptive regulatory tools but also as ongoing monitoring points, allowing authorities to review the underlying algorithms and operational security of digital virtual human services. The public comment period itself, ending on May 6, 2026, represents an initial phase of evaluation, gathering feedback from industry, experts, and the public to refine the regulations. This iterative approach to monitoring and evaluation aims to build a responsive regulatory system that can effectively address new challenges and foster innovation while upholding safety and ethical standards in the rapidly evolving field of AI-generated digital humans.

Penalties, Liability, and Appeals

While the detailed provisions for penalties, liability, and appeals are typically elaborated in the final version of a regulation, the Draft Measures lay a foundational framework for accountability. The inclusion of a chapter on "Supervision, Inspection and Legal Liability" (监督检查与法律责任) clearly indicates the intent to establish a robust enforcement mechanism. This chapter outlines the legal consequences for non-compliance with the various provisions set forth in the Draft Measures. Violations are expected to trigger administrative penalties, which could range from warnings and orders for rectification to fines, suspension of services, or even revocation of licenses, depending on the severity and nature of the infringement. The multi-agency oversight, involving departments like public security and market regulation, suggests that enforcement actions could be coordinated and comprehensive, reflecting the broad impact of digital virtual human services.

Beyond administrative penalties, the Draft Measures also imply avenues for civil liability and potential criminal accountability where applicable. For instance, infringements on personal rights such as reputation, privacy, and intellectual property, as well as the misuse of personal information without explicit consent, could lead to civil lawsuits for damages. The prohibition against using digital humans to bypass identity authentication or engage in fraudulent activities further suggests that severe violations could fall under existing laws related to cybersecurity, fraud, or other criminal offenses. While specific appeal mechanisms are not detailed in the available snippets, it is generally understood within the Chinese legal system that entities subject to administrative penalties have rights to administrative reconsideration or administrative litigation to challenge regulatory decisions. The emphasis on protecting the legitimate rights and interests of individuals and organizations throughout the Draft Measures underscores the importance of effective redress mechanisms for those adversely affected by non-compliant digital virtual human services.

Relationship to Other Instruments

The Draft Measures are not developed in isolation but are intricately woven into China's existing and evolving legal and regulatory landscape concerning cyberspace, data, and artificial intelligence. The foundational legal basis for these Draft Measures is explicitly stated to include several key national laws and administrative regulations. These include the "中华人民共和国网络安全法" (Cybersecurity Law of the People's Republic of China), the "中华人民共和国数据安全法" (Data Security Law of the People's Republic of China), and the "中华人民共和国个人信息保护法" (Personal Information Protection Law of the People's Republic of China). These overarching laws provide the fundamental principles and requirements for network security, data protection, and individual privacy, to which the digital virtual human regulations must adhere and further elaborate upon.

Furthermore, the Draft Measures build upon and complement other specific regulations, such as the "互联网信息服务管理办法" (Administrative Measures for Internet Information Services), the "未成年人网络保护条例" (Regulations on Online Protection of Minors), and the "网络数据安全管理条例" (Regulations on Network Data Security Management). Notably, the Draft Measures are also designed to align with and extend the principles established in the Cyberspace Administration of China's (CAC) 2025 Measures for Labeling of AI-Generated Synthetic Content, particularly regarding transparency and content identification. For digital virtual human services with public opinion or social mobilization capabilities, the Draft Measures reference the requirements of the "互联网信息服务算法推荐管理规定" (Internet Information Service Algorithm Recommendation Management Provisions) for algorithm filing and security assessments. This interconnectedness demonstrates a strategic effort to create a coherent and comprehensive regulatory ecosystem for AI, ensuring consistency and avoiding fragmentation across different technological domains and policy objectives.

International Alignment

The Draft Measures, while primarily focused on domestic regulation, also reflect China's awareness of and engagement with international discussions and best practices in AI governance. The document itself emphasizes a balanced approach of "development and security" and encourages "standard-setting and international cooperation." This indicates an intent to not only regulate internally but also to contribute to and potentially align with global efforts in establishing norms and standards for responsible AI development and deployment. The emphasis on "AI for good" and a "people-oriented" approach resonates with ethical AI principles being discussed and adopted by various international bodies and countries worldwide.

While the Draft Measures are a national initiative, their broad scope and detailed provisions on issues such as data protection, transparency, and the protection of vulnerable groups (like minors) touch upon universal concerns in AI governance. By addressing these issues comprehensively, China's regulatory framework for digital virtual humans could potentially serve as a reference point or contribute to the evolving global discourse on AI ethics and regulation. The encouragement for active participation in international rule-making and exchanges suggests that China views its regulatory development not in isolation but as part of a larger global effort to navigate the complexities of advanced AI technologies. This proactive stance positions China as a significant player in shaping the future of international AI governance.

Implementation Timeline

Sources and References