Anna is here. I opened Claude last Tuesday and got a prompt I wasn't expecting: a "quick identity check" asking for a government ID and a selfie. I stared at it for a minute. I'd been using Claude almost every day for months. Nothing had changed on my end. Why now?
If you're here, you probably hit the same wall. This isn't a piece about whether Anthropic should be doing this — there's plenty of that elsewhere. This is what I figured out after digging through their help center, reading the coverage, and trying to decide what I actually wanted to do about it.
Why you're suddenly being asked
What triggered the prompt (from what's known)
Anthropic hasn't published a full list of triggers, and I want to be honest about that upfront — anyone telling you they know exactly why your account got flagged is guessing. What the official help center page does say is that verification shows up "when accessing certain capabilities, as part of our routine platform integrity checks, or other safety and compliance measures."
That's vague on purpose, I think. But pulling from the reporting, a few patterns seem to come up:
Accounts that look like they might be under 18
Accounts showing signs of being in unsupported regions — mainland China, Russia, Iran, and a few others
Patterns Anthropic's classifiers read as potential policy violations
Some higher-tier or more sensitive capability access
Why it didn't happen last month
Anthropic quietly rolled out the verification requirement in mid-April 2026. The help center page appeared around April 14–16, and users started seeing prompts shortly after. So if you used Claude fine in March and got hit with this in April, you're not imagining it. The system is new.
What the verification actually does
Documents that work and documents that don't
The accepted list is narrow. You need a physical, original, government-issued photo ID. That means:
A passport
A driver's license or state/provincial ID
A national identity card
What doesn't count:
Photocopies, scans, or screenshots
Digital/mobile driver's licenses (even official ones)
Student IDs, work badges, bank cards, library cards
Temporary paper IDs
Expired documents
The document also has to be undamaged and clearly readable. A creased passport or a license with a worn photo might fail even if it's technically valid.
The selfie step and what it compares against
You'll likely also be asked for a live selfie through your phone camera or webcam. The point is to confirm the person holding the ID is the one in the photo — a standard liveness + face match check. Anthropic uses Persona as its verification partner, the same infrastructure many financial services use for KYC.
Persona holds the ID and selfie on their systems, not Anthropic's. That's worth knowing either way, because it means your documents pass through a third party.
What happens if you decline
Feature loss vs full account restriction
This is the part I wanted to understand most, and the honest answer is that it depends on why you're being asked. If you're being verified as part of a capability check, skipping it likely means you lose access to that specific capability. If you're being verified because the system flagged your account for a potential policy issue — age, region, usage — refusing to verify can lead to the account being paused or banned.
The help center is careful to list bans as a separate category (reasons include repeated usage policy violations, unsupported location, ToS violations, and under-18 usage) rather than framing "declined to verify" as an automatic ban trigger. But practically, if you can't or won't verify, and the system already thinks there's a compliance reason to check, you're probably going to lose access.
Appeals and retries
If your verification fails — blurry photo, bad lighting, ID not reading cleanly — you get multiple attempts inside the same flow. Most failures, based on what Anthropic says, are fixable by redoing the photo or switching to a different accepted ID.
If you've used all your attempts and still can't get through, there's a support form for verification-specific issues. And if your account gets banned and you think it's a mistake, there's a separate appeal form linked from the same help center page. I've seen reports of accounts being restored after appeal, though often with delays and lost project history in the interim.
Privacy steps you can take either way
Reviewing the Persona data flow
Whether you verify or not, it's worth understanding where your data goes. Anthropic is the data controller, Persona is the processor. Both say ID images aren't used to train models and aren't shared for marketing. Data is encrypted in transit and at rest.
The gap worth noting: Anthropic says it sets the retention period but, as The Register pointed out, hasn't publicly stated what that period is. If that matters to you, the privacy policy is where the details would eventually surface.
Account hygiene recommendations
A few things I did that I'd recommend regardless:
Export anything you care about from existing chats and Projects. If something goes sideways with verification or a flag, you don't want to lose work while waiting on an appeal.
Double-check the email on your account. Appeals go through there.
If you're paying for Pro, make a mental note of when your billing cycle ends. Some users who got suspended have been refunded for the current cycle, but that's not guaranteed.
Alternatives if you'd rather not verify
Scaling back to free-tier capabilities
I don't know if this will work for everyone — again, triggers aren't fully public — but if the verification prompt only shows up for certain capabilities, you may still be able to use Claude in a more limited way without verifying. This is worth testing before you decide.
Moving your workflow to other AI tools
As of right now, neither ChatGPT nor Gemini requires government ID verification for standard consumer use. Whether that stays true is another question — age verification pressure is increasing across the whole category, and OpenAI has been testing age-estimation systems. But if verification is a firm no for you, those are the obvious alternatives, and most workflows port reasonably well.
I'll say this honestly: I'm not going to tell you which tool is "better." They're different, and which one fits depends on what you actually use Claude for. The verification question doesn't change that — it just changes the cost of staying.
FAQ
Can I use Claude without a passport?
Yes, if you have another accepted government-issued photo ID — a driver's license, state ID, or national identity card. Passports aren't specifically required; any physical, original, government-issued photo ID that clearly shows your face works.
Does failing verification ban my account?
Not automatically. Verification failures — blurry photos, unreadable IDs — give you multiple retry attempts, and you can contact support if you run out. A ban is a separate thing, tied to reasons like repeated usage policy violations, unsupported regions, or being flagged as under 18. The two can overlap, but they aren't the same process.
Is my selfie used to train Claude?
No. Anthropic states explicitly that verification data — ID and selfie — is not used to train models. It's used only to confirm identity and meet legal and safety obligations. The data sits with Persona, not on Anthropic's systems.
Can I delete my verification data later?
Anthropic says data is deleted according to retention limits they've set with Persona and applicable law, but hasn't publicly stated what those retention periods are. If deletion matters to you, the support form or privacy policy contact is where to ask for specifics on your case.
That's where I am with it. I haven't decided yet whether I'll verify — I keep going back and forth. It's a small friction, and also not a small friction, depending on the day.
If you're in the same spot, you're not overreacting. You're just paying attention.
Hi, I'm Anna, an AI exploration blogger!
After three years in the workforce, I caught the AI wave—it transformed my job and daily life. While it brought endless convenience, it also kept me constantly learning. As someone who loves exploring and sharing, I use AI to streamline tasks and projects: I tap into it to organize routines, test surprises, or deal with mishaps. If you're riding this wave too, join me in exploring and discovering more fun!
