Anthropic Now Wants Your Passport. Users Are Leaving.

Key Takeaways - Anthropic's ID verification kicks in July 8, 2026. Government photo ID, live selfie, and facial geometry scan required for consumer accounts. - Fable 5 and Mythos 5 vanished in hours after a reported jailbreak, proving your AI access can disappear overnight. - API users are exempt. Team, Enterprise, and business accounts aren't subject to the July 8 consumer requirement. - Mistral + Ollama setup takes about two hours. That's your window before the deadline hits.

---

Anthropic updated its privacy policy on June 15, 2026. Starting July 8, consumer account holders on Claude Free, Pro. And Max can be asked to hand over a government-issued photo ID, a live selfie, and a facial geometry scan.

Refuse and Anthropic can suspend your access.

This isn't theoretical.

It's already happening to some users.

The backlash on Hacker News hit 754 points.

Fireship put out a video titled "One man just liberated Fable... and now it's illegal." Non-US users are migrating to Mistral. This is less than a week old and still raw.

Anthropic ID Verification Policy: What Changed

The short version: the US government forced its hand.

Fable 5 and Mythos 5 went dark worldwide on June 12 after a reported jailbreak, with no public explanation of the national security rationale. Anthropic complied. Days later, it began enforcing government ID verification through Persona for certain Claude capabilities.

The policy update extending this to all consumer accounts isn't just a response to the jailbreak. It's a broader shift toward treating AI access the way banks treat account holders.

Anthropic says the ID checks are voluntary.

You can refuse, but you may lose access. The policy doesn't spell out what happens if you say no. The Facebook explainer making the rounds says refusal leads to suspension. The Register notes that different jurisdictions may require other responses, which explains the vague language.

Either way, the direction is clear: if you want to keep using Claude as a consumer, you may need to verify who you are.

The timing matters. Fable 5 and Mythos 5 vanished in hours. ID verification arrived days later.

This is what platform dependency looks like when a regulator decides it has a concern.

Your access can vanish between one update and the next.

What Claude Users Must Submit

If Anthropic asks you to verify, the requirements are specific:

- Physical passport - Driver's license - State ID

Here's what doesn't count: photocopies, screenshots, photos of photos, digital mobile IDs, employee badges.

The live selfie must be clear and undamaged. Persona collects and holds the images. Anthropic says it doesn't store the raw photos on its own systems.

That's the official line.

The data Anthropic can collect under the updated policy is broader than the current implementation. Starting July 8, it may collect:

- An image of your government ID including ID number and date of birth - Your image in photo or video form - Facial geometry templates (which may constitute biometric data) - The verification result itself. Whether you passed or failed the age threshold

The policy also includes discretionary language about sharing data with authorities "in certain circumstances," which The Register flagged as substantially broader than the previous version.

The policy explicitly excludes commercial customers.

Team, Enterprise, and API accounts are not subject to the July 8 consumer verification requirement. If you're building on Claude as part of a business, this is a meaningful distinction.

Though the practical friction for individual developers is already real.

Why Privacy-Conscious Users Feel Betrayed

Anthropic built part of its brand on being the privacy-conscious alternative to OpenAI.

It targeted users who didn't want their data harvested. One Times of India reader put it bluntly: they switched to Claude specifically because of concerns about OpenAI's surveillance practices. And now Anthropic is asking for a government ID.

The response on social media has been sharp.

"Anthropic just handed ChatGPT and Gemini a gift," one X user wrote. Another said Anthropic "collapsed the boundary between identity and thought." A Hacker News thread attracted 623 comments, with developers calling the move "surveillance" and asking why a reasoning model needs a facial geometry template.

The developer I keep coming back to is the one who switched from Claude to Gemini last month after being asked for ID. Not as they did anything wrong. But since they didn't want to hand over a government document for the privilege of using an AI chat interface. That sentiment is real and it's spreading.

How to Prepare Before July 8

Anthropic will likely add more verification triggers as the policy matures.

The Fable and Mythos shutdown shows how fast access can disappear. For smaller operators and indie developers, the lesson is uncomfortable: any single AI provider can change terms overnight. The ID verification isn't the risk. The assumption that it won't affect you is.

Here's what I did on my own workload this week:

1. Set up a Mistral account 2. Install Ollama on a spare machine 3. Move non-production prompts to the alternative provider 4. Test switching times

Took about two hours. If Anthropic asks me to upload a passport, I want a working fallback before I make that call.

This applies to anyone running production work on a single LLM provider, not just Claude users worried about this specific policy. The ID verification concern is the trigger, but the underlying issue is vendor dependency. When the government decides an AI model poses a national security risk. As it did with Fable 5. Your access is collateral.

You don't get a vote.

FAQ: Anthropic ID Verification

Is Anthropic requiring ID to use Claude?

Yes, for consumer accounts (Free, Pro, Max) starting July 8, 2026. Users can be asked to submit a government-issued photo ID, live selfie, and facial geometry scan or risk suspension.

Can I refuse the Claude ID check?

You can refuse, but Anthropic may suspend your access. The policy language is vague on exact consequences, but the direction is clear: refusal likely means losing access.

Does this apply to API users?

No. Team, Enterprise, and API accounts are explicitly excluded from the July 8 consumer verification requirement.

What are alternatives to Claude?

Mistral (via API or webchat) and Ollama (local self-hosted) are the most direct alternatives mentioned by developers making the switch.

Setting up a fallback takes roughly two hours.

The Bigger Picture

Anthropic is caught between two pressures. The US government is treating frontier AI as a strategic asset that requires access controls. Users chose Claude in part since it didn't operate that way. Anthropic can't fully satisfy both, and right now it's choosing government compliance over user trust.

That calculation may be correct. But it's the moment competitors use to pull users away.

The July 8 deadline is a useful forcing function. Either Anthropic walks this back before it fully takes effect, or developers make the switch. For smaller operators, the move should be obvious: configure a second provider before you need it, not after you've already lost access.

Your AI stack shouldn't be a single point of failure. Not as AI is fragile, but since governments are getting more interested in controlling who uses it and why.

The question for you is simple. Do you want your AI access gated by a government ID check? If not, you have until July 8 to do something about it.

Sources

- Anthropic Privacy Policy Update — June 15, 2026 - The Register. Anthropic's broader data-sharing language flagged - Hacker News, 754-point thread on Claude ID verification - Fireship — "One man just liberated Fable... and now it's illegal" - Times of India — User switching from OpenAI to Claude over surveillance concerns