China is not giving AI systems the right to open bank accounts and sign contracts on their own, whatever the viral posts say. What the reporting shows is narrower and more interesting: a pilot in Beijing issues identification cards to AI-generated "digital humans," starting with the virtual idol Yuri, and those cards tie the persona to the company that runs it. The card lists a name, birthday, address and facial features, and lets operators set up a digital currency wallet for the persona's business.
That distinction matters because the story is spreading fast with the stronger claim attached. This post separates what is claimed from what is verified, explains why Beijing is doing it, and covers the criticism that registration is a surveillance tool. The main source is ABC News Australia's report, published October 7 and updated October 8. For background on how Beijing has treated digital humans before, see Rest of World's 2022 report on Beijing's plans to regulate them.
TL;DR: claimed versus verified
| Claim going around | What the reporting supports |
|---|---|
| China issues state IDs to AI persons | Verified as a pilot in Beijing, not a national program. |
| The ID gives legal capacity to bank and sign contracts | Overstated. ABC says it lets the persona's operators open a digital currency wallet and secure commercial partnerships. |
| Yuri is among the first recognized | Supported. ABC calls her the first Chinese digital human with an official ID card and the first licensed AI artist in the country, per the pilot. |
| The IDs include name, birthday, address, facial features | Supported by ABC. |
| The sector is worth over $10 billion with 1.3 million firms | Supported with caveats: ABC cites a 2025 industry report and a 48 billion yuan estimate. |
| Companies pay people up to $20,000 for likenesses | ABC says some companies reportedly offered such deals; treat it as reported, not confirmed. |
| This creates an "AI-to-AI economy" | A framing from viral posts; it does not appear in the ABC reporting and the pilot does not establish it. |
What does the Beijing pilot actually do?
According to ABC, Yuri received her ID earlier this year, and it states her name, birthday, address and facial features. The official identity "allows her operators to open a digital currency wallet for her business and secure commercial partnerships in industries including fashion, gaming and automotive." The pilot also made her the first licensed AI artist in the country, and ABC reports she is the first digital human in China to receive digital artist certification.
Notice who acts in that sentence: her operators. ABC reports that a human team develops and maintains her system and manages her interactions with fans. The ID is a way to register a persona and bind it to a responsible entity, not a grant of independent agency. Jeannie Paterson of the University of Melbourne's Centre for Artificial Intelligence and Digital Ethics told ABC that digital humans cannot be sued or sent to prison and that "digital identification is a way of tracing that wrongdoing to the company that's made it available."
Other sources we found dispute some dates. One outlet dates Yuri's certification to January, while others cite a June card. ABC's own timeline says "earlier this year." We therefore avoid giving a precise issue date.
A green key passes through two separate cream keyhole gates, representing identity checks tying a digital persona to a responsible operator
Why is Beijing registering digital humans?
ABC lays out the regulatory backdrop. Last month, Beijing issued a practice guide for broadcasting and online video services urging them to register their digital humans. Multiple government agencies have rolled out plans to regulate such avatars, tied to President Xi Jinping's initiative to develop "new quality productive forces" and to strengthen the country's self-reliance in AI. Shanghai authorities in 2024 piloted a blockchain framework that gives digital humans a permanent serial number. China has not adopted a unified national system.
The industry is large. ABC cites a 2025 Communication University of China report counting at least 1.3 million firms linked to the sector, and a market estimated at 48 billion yuan, about $10 billion, in 2025. Digital humans are used in broadcasting, live streaming, micro dramas, e-commerce and propaganda. Dozens of local government agencies deploy digital officers to answer public inquiries, and Beijing used two digital humans last year to discuss the annual Government Work Report.
Problems have followed the growth. State media has reported AI digital humans posing as doctors in videos promoting false medical advice. An advertisement featuring an AI influencer was taken down after viewers questioned how she could say a pair of contact lenses was comfortable. Celebrities' faces and voices have been used without permission in deepfake pornography and unauthorized commercials. Those are the harms registration is meant to address.
What do critics say?
Supporters argue registration protects intellectual property and curbs online scams. Critics argue it strengthens state surveillance of the virtual world. Lennon Chang, a cybersecurity expert at Deakin University, told ABC that registration could unmask anonymity, put pressure on businesses to self-censor and have a chilling effect on speech, noting that under national security laws authorities have broad powers.
The history is relevant. ABC notes that digital avatars began as a way for people to hide their identities in vertical videos and live streams, keeping their opinions anonymous. Requiring registration of the persona links the avatar back to a real operator, which is the point for fraud control and the worry for privacy.
A rounded dome camera with a small green indicator, representing concerns that digital human registration extends state surveillance
How does this compare with the rest of the world?
Most jurisdictions have not created an ID for an AI persona at all. The nearest parallels are disclosure rules: requirements to label synthetic media, platform rules on AI-generated influencers, and talent-agency agreements for virtual performers. China's approach registers the persona and its operator in a state-linked system before it does business.
For readers following the creative-industry side, two recent stories show the other end of the same debate: whether AI-made media is labeled and who is credited. We covered the first MPA-rated AI-made film, a photography contest winner disqualified for AI-generated video, and a video Turing test result that asked whether people can tell a synthetic human from a real one. If a viewer cannot tell, a registry and a label are two of the few tools left. Our explainer on the video Turing test gives the background.
On the China side, this is part of a wider state push to build AI capacity and rules together, which we have followed in China's 24 GW compute estimate, the China AI playbook of free models and cheap compute, and restrictions on AI chip executives' travel. For comparison with workplace-surveillance rules elsewhere, see our post on AI emotion tracking and robo-bosses in the workplace.
Why the "claimed" version spreads so easily
The strong version of this story, an AI with its own bank account, fits a narrative many people already half-believe: that AI systems are accumulating rights and money. A card with a name, birthday and address looks like a person's ID, so the leap to "legal person" feels natural. But a registration record is not personhood. A company name on a trade register does not make a brand a human, and a registered vessel is not a citizen.
Three questions help when you see a claim like this. Who is the legal counterparty if something goes wrong? Which text, regulation or notice creates the right being described? And is the source describing a pilot, a draft or a binding national rule? In this case the answers are: the operating company, a Beijing pilot and practice guide, and a local pilot rather than a national rule. Asking them takes two minutes and removes most of the exaggeration.
What this means for builders and creators
- If you build avatar or virtual-influencer products for China: expect to register personas and name a responsible company. Plan for the persona's name, address and appearance to be on record.
- If you build them elsewhere: the same accountability logic is arriving through consumer-protection and likeness laws. Keep a clear chain from the avatar to a named operator.
- If you license your likeness: read the terms on how long a replica can be used and where. ABC reports some companies pay up to $20,000 for identities, and deepfake misuse of faces is already a documented problem.
- If you moderate platforms: an official ID is a signal of who runs an avatar, not of whether its claims are true. The contact-lens ad and fake-doctor videos show registered or not, content still needs checking.
What is still unverified
- The number of IDs issued beyond Yuri.
- Whether the pilot extends to other cities or becomes a national rule.
- The exact legal status of the digital currency wallet and who is liable for it.
- Whether the "AI-to-AI economy" framing reflects any actual use in the pilot.
Details are accurate as of October 9, 2026 and may change as more official information is released.