Claude, forced to "smoke your face and check your accounts" and no identity cards since July

Author:Shinjimoto

 

Anthropic, another big deal

In recent days, several Claude users have opened the mailbox and found an email from Anthropic. This post is part of our special coverage “Personal Policy Update”. But the e-mail content surprised them:

"We might ask you to verify age or identity."

SOMEONE TANNED AN E-MAIL SCREEN ON X, AND THE COMMENT AREA EXPLODED IN A FLASH. ACCORDING TO THE EMAILStarted on 8 JulyClaude's probably gonna need your ID. Anthropic's real name is finally coming

Anthropic suddenly announced that the real name was coming

June 10thMany Claude users received emails from Anthropic in their mailboxes。

In the mail, from8 JulyFrom now on, Anthropic may require users to verify age or identity on the grounds of "enhanced security". More interestingly, it was discovered that they had just received the e-mail today, but other accounts had previously been confiscated, presumably in batches and greyscale tests。

According to the official updated privacy policy in the mail, three core changes were made:

1. Multi-step task and third-party data flow

As Claude's ability is enhanced, it can perform multi-step missions on behalf of users (e.g. plan a full trip for you: check flights, book hotels, schedule trips) or connect to third-party applications (e.g. Google Drive, Slack, Notion, etc.)。

Besides, your data is no longer just on the Anthropic server. In its updated privacy policy, Anthropic states:When you connect a service or make Claude represent you on a mission, data can flow between you and a third party。

This means that your conversation content, uploading files, and even your operating instructions may leave the Anthropic server for third-party applications. For ordinary users, the stronger the "agent power" of AI, the wider your data footprint。

Validation of data (core change)

This is the most powerful element of this update. The policy reads:

As one of our measures to ensure the safety and reliability of services, we may require you to verify age or identity. We are here to explain what information we will gather and how it will be collected。

How

Validation via third party service according to Anthropic Official Support PagePersonaGo ahead. Specific processes include:

• Uploading Government-issued identity documents with photographs (passports, driver ' s licences, identity cards, etc.)

• Take a real-time self-portrait (for matching identity photographs)

Anthropic tries to appease the users by emphasizing that the data are used only for identification purposesNot for model training; data will not be stored on Anthropic's own server (processed by Persona); and the entire process will only target specific security and compliance scenarios。

However, the announcement continues to cause a pupil earthquake among a wide range of users. On the X platform, a user tanned an e-mail screen with text: "On June 10, it was received and thought it was an ordinary update ... Now look back and think very carefully. And the user's play: They said they wanted to authenticate, so Fable 5 is coming back

3. Research participation information

If you attend anthropic research, interviews, they collect more information。

Focus:

3 changes above, only forConsumer level accountsFree, Pro, Max, commercial clients (Team, Enterprise) are not affected. In other words, ordinary individual users will be the first "validated" targets。

Anthropic explains on the official support page that identification was previously used only in specific scenarios targeting minors, irregular users or regional accounts. In the future, with the complexity of proxy tasks, the scope of certification will be gradually expanded。

Actually, this ID thing, Anthropic, it's not the first time. Similar means of verification had previously been used in dealing with the protection of minors, irregular accounts and accounts in partially unsupported areas. The difference is that the wording and scope are clearer this time, and the sentence is deliberately drawn —The validation scene may continue to expand as the Agent mission becomes more complex in the future。

Many Agent users are worried: will you be asked to brush your face with a window if you can't move? There are currently no more detailed trigger conditions, and the rate will be more specific after July 8。

"THE WILD AGE" BY AI IS COMING TO AN END

The recent "Fable 5 Sealed" events seem to have premonitioned。

The original Fable 5 account was officially blocked for touching the security red line and shocked everyone. Many believe that this may be a disciplinary action against individual extreme cases, but it now seems that the fall of Fable 5 is not an isolated incident, but rather a prelude to the full tightening of the Anthropic and even the AI industry as a whole。

When AI’s ability evolved from a “talking tool” to an “all-power agent” capable of operating a complex third-party application, the official will never allow the next Fable 5 to continue running outside the law。

The so-called "real name" and "facewashing" markOn the eve of the full-scale outbreak of AI Agent, the giants must put in place an indestructible accountability mechanism ahead of schedule。ONCE AI HAS MADE A MESS IN EXECUTING BOOKINGS, AMENDING DOCUMENTS AND EVEN MORE COMPLEX BUSINESS DIRECTIVES ON YOUR BEHALF, THE PLATFORM MUST KNOW EXACTLY WHICH "PERSON" IS REALLY BEHIND THE SCREEN。

From8 JulyAt first, every time you try to get Claude to help you with the complex task of multi-steps, you might be accompanied by a bullet window: upload your identity papers。

Anthropic revealed that winter is coming. And from that point on, brushing, sealing, probably becoming daily in the AI industry in 2026。

In a month, when Claude is in use, we'll be ready to raise our ID。

References:

• https://x.com/ryanmcadams/status/2066201356034011370?s=20

• https://privacy.claude.com/en/articles/103301952-updates-to-our-privacy-polic