Claude enforces "facial recognition for household registration," starting in July, no ID card means no access?

Core Viewpoint

Anthropic suddenly announces, is real-name verification really coming?

On June 10, many Claude users received an email from Anthropic.

The email mentioned that starting from July 8, Anthropic may require users to verify their age or identity for "enhanced security." Interestingly, some users found that they just received this email today, while other accounts had not received it before, suggesting a staggered rollout or gray testing.

According to the updated privacy policy mentioned in the email, there are three core changes:

1. Data flow of multi-step tasks and third-party applications

As Claude's capabilities improve, it can perform multi-step tasks on behalf of users (such as planning a complete trip: checking flights, booking hotels, arranging itineraries) and can connect to third-party applications (such as Google Drive, Slack, Notion, etc.).

Moreover, your data will no longer just stay on Anthropic's servers. Anthropic clearly states in the updated privacy policy: When you connect to a service or let Claude perform tasks on your behalf, data can flow between you and third parties.

This means that your conversation content, uploaded files, and even your operational commands may leave Anthropic's servers and enter the processing flow of third-party applications. For ordinary users, the stronger the AI's "agent capabilities," the wider your data footprint will be.

2. Verification data (core change)

This is the most impactful content in this update. The policy states:

As one of our measures to ensure the security and reliability of our services, we may ask you to verify your age or identity. We hereby explain what information we will collect and how we will collect it.

How to verify?

According to Anthropic's official support page, verification will be conducted through a third-party service called Persona. The specific process includes:

Uploading a government-issued photo ID (passport, driver's license, ID card, etc.)
Taking a real-time selfie (to compare with the ID photo)

Anthropic attempts to reassure users, emphasizing that this data is only used to confirm identity and will not be used for model training; the data will not be stored on Anthropic's own servers (processed by Persona); and the entire process is only for specific security and compliance scenarios.

However, this notification still left many users in shock. On the X platform, one user shared a screenshot of the email with the text: "Received on June 10, thought it was just a regular update… looking back now, it's quite terrifying." Another user jokingly remarked: They said they would verify identities, so Fable 5 is making a comeback?

3. Research participation information

If you participate in Anthropic's research or interviews, they will collect more information.

⚠️ Key note:

The above three changes apply only to consumer accounts (Free, Pro, Max), and commercial clients (Team, Enterprise) are not affected. This means that ordinary individual users will be the first group to be "verified."

Anthropic explains on its official support page that identity verification was previously only used in specific scenarios for minors, violators, or accounts from unsupported regions. In the future, as agent tasks become more complex, the scope of verification will gradually expand.

In fact, this is not the first time Anthropic has implemented identity verification. Similar verification methods have been used previously for handling minors' protection, violating accounts, and some accounts from unsupported regions. The difference this time is that the wording and scope are clearer, and it specifically points out that—as agent tasks become increasingly complex, the verification scenarios may continue to expand.

Many heavy users of agents are concerned: in the future, when using Claude for automated tasks, will they frequently be prompted to verify their identity? Currently, the official has not provided more detailed triggering conditions, but after July 8, there will likely be more specific explanations.

The "Wild Era" of AI is coming to an end

Looking back at the recent uproar over the "Fable 5 being banned" incident, everything seems to have been foreshadowed.

Initially, the Fable 5 account was banned by the official due to crossing security red lines, shocking everyone. Many thought this might be a punishment for a few extreme cases, but now it seems that the downfall of Fable 5 is by no means an isolated incident, but rather a prelude to Anthropic and the entire AI industry tightening regulations.

As AI's capabilities evolve from being a "chat companion" to an "omnipotent agent" capable of operating third-party applications and executing complex tasks, the authorities will not allow the next Fable 5 to continue running rampant in a legal gray area.

The so-called "real-name system" and "facial recognition verification" signify that on the eve of the full outbreak of AI agents, major players must establish an unbreakable accountability mechanism in advance. Once AI encounters issues while executing tasks such as booking tickets, modifying documents, or even more complex business instructions on your behalf, the platform must know exactly which real "person" is giving the orders behind the screen.

Starting from July 8, every time you attempt to have Claude assist you with complex multi-step tasks, it may be accompanied by a prompt: please upload your ID.

Anthropic's message reveals—winter is coming. From now on, facial recognition verification and account bans will likely become the norm in the AI industry by 2026.

In a month, when using Claude, be prepared to show your ID at any time.

References: