Show Your ID to Use Claude? Anthropic's Quiet Push Toward Identity Verification

The days of logging into an AI chatbot with nothing but an email and password are ending. Slowly, quietly, the prompts are changing: hold your ID up to the camera, take a selfie so we can match your face. Anthropic, the company behind Claude, is at the center of this shift. And a question is starting to surface that the industry has mostly avoided: why does using an AI require proving who you are?

Let me be straight about where this stands. This isn’t a topic blowing up in the community yet. Over the past month, there’s been no firestorm on Reddit or Hacker News. The clearest signal so far is a YouTube video titled “The End of Anonymity, Claude’s Identity Ultimatum” — a spark catching underground rather than a blaze. So instead of chasing facts that aren’t fully there, let’s dig into the structural question: why is this happening at all?

Why ID Verification Is Suddenly Knocking on AI’s Door

Identity checks didn’t fall out of the sky. Three pressures are pushing in the same direction at once.

The first is regulation. Governments everywhere are trying to put distance between AI and minors. Norway recently moved to effectively ban AI in elementary schools. But enforcing protections for minors requires answering one question — is this user an adult? — and the most airtight way to verify age is a government ID.

The second is abuse prevention. A powerful model is a double-edged sword. Anonymous users with unlimited access open the door to mass spam, fraud, and coordinated influence operations. Tie one account to one real name, and the cost of abuse shoots up.

The third is accountability. When AI causes harm, someone has to answer for it. If you don’t know who the user is, you can’t trace them or block them. For a company, identity verification is a kind of insurance policy.

“Safety” as the Pitch, Surveillance as the Shadow

Here’s where the core tension lives. The justification is always safety — protect minors, filter out bad actors. It’s a hard intention to argue against.

The problem is the side effect that good intention creates. Submit your ID once, and every question you’ve ever asked Claude becomes permanently linked to a single, named human. What you’ve worried about, what illness you’ve quietly searched, what political question you typed at 2 a.m. — none of it is anonymous anymore.

That YouTube title — “The End of Anonymity” — isn’t hyperbole. Even in the search-engine era, there were at least token efforts to decouple queries from real identities. But AI conversations are far more intimate than search. People tell chatbots things they wouldn’t tell a friend or a doctor. The moment that data is bound to a legal name, its weight changes entirely.

Who Gets to Use AI?

Step back and the frame widens. ID requirements aren’t just a privacy issue — they’re an access issue.

Plenty of people can’t, or won’t, hand over a government ID. People in countries where getting one is difficult. People fleeing political persecution. And ordinary people who simply don’t want to surrender their identity to a trillion-dollar company. If verification becomes the norm, all of them are effectively locked out of the most capable AI tools.

As AI becomes core infrastructure for working, learning, and living, deciding who’s allowed onto that infrastructure becomes a form of power. A rule that says “only those who can prove their identity get to use AI” can build a digital class system — regardless of anyone’s intent.

What’s Actually Worth Watching

Anthropic hasn’t mandated ID for every user. Right now it looks limited to specific situations — higher-risk capabilities, certain regions. But remember the pattern: once one player sets the standard, the rest of the industry tends to follow.

Watch three things. First, scope: does verification stay confined to a handful of high-risk features, or does it become a condition of logging in at all? That’s the watershed. Second, method: does the company route checks through a third-party verifier instead of holding your documents itself, and does it delete the data immediately after? That’s what separates trust from theater. Third, choice: is there still a door — even a narrower one — left open for users who decline to verify?

Nobody’s against making AI safer. The question is whether anonymity and access are the right price to pay for it. So where do you land? Would you hand over your ID for a smarter Claude — or is this one of those lines we shouldn’t give up so easily?