Anthropic May Ask Claude Users to Verify ID

According to Anthropic spokesperson Michael Aciman, the change applies only to a small subset of users whose accounts are flagged but not outright banned. A post from Anthropic Thariq Shihipar said the policy was updated on June 17 as an update to the appeals process and is unrelated to the Fable or Mythos rollout. Anthropic would not say how many people a subset represents, though the company is thought to have tens of millions of monthly users.

How the identity check works mechanically

When the check is triggered, the flow is concrete and biometric-heavy. The policy would require the affected user to upload a photo scan of a government-issued passport or driver license. Anthropic says it will also collect a selfie photo or video and a digitized version of the person as a face geometry template.

That last item matters. Some states, like Illinois, legally classify face geometry templates as protected biometric data, which carries its own consent and storage obligations. Anthropic also says it keeps a record of the verification result, such as whether the user has reached a certain age.

The work is not done in-house. Anthropic says it uses a San Francisco company called Persona as its identity checking provider. Users may see a verification prompt when accessing certain capabilities, as part of routine platform integrity checks, or under other safety and compliance measures.

If you are asked to verify, confirm the prompt is genuine before uploading anything. A government ID, a selfie, and a face scan are exactly the data a phishing page would want, so only complete verification inside the official Claude interface and never via a link sent over email or chat.

The data-retention question Anthropic did not fully answer

The biggest open issue is what happens to your documents afterward. Anthropic says it decides how long Persona retains user identity documents, but its spokesperson did not immediately say when that data is deleted.

That silence stands out against the alternative. Roblox, another Persona customer, says user images are deleted immediately after processing, which limits the window in which the information could be leaked or stolen. Without a stated deletion timeline, Claude users are left trusting that the data does not linger.

There is also a jurisdictional wrinkle. Persona can still face U.S. government demands for the user information it stores on its servers. So even if Anthropic never directly hands over your ID, the vendor holding it could be compelled to.

Persona itself is not a neutral detail. It is backed by Founders Fund, the investment firm founded by Trump backer Peter Thiel, who also invests in Anthropic. The link has drawn criticism before. Earlier this year Discord selected Persona for age verification, then quickly reversed course after user backlash over the Thiel connection.

Why this is happening now

The timing reads as more than housekeeping. Keeping closer tabs on who is using its tools may help Anthropic navigate a stack of legal challenges, regulatory changes, and political pressure.

The company remains largely at an impasse with the White House. More than a week earlier, Trump officials effectively forced Anthropic to pull its latest cybersecurity models over allegations that a jailbreak could break the models guardrails. Other reporting has pointed to personality clashes as the deeper cause of the breakdown. Months before that, the Department of Defense designated Anthropic a supply chain risk, reportedly in retaliation for refusing to let the government use its technology for mass domestic surveillance or to power fully autonomous weapons.

What happens next in the coming 24 to 72 hours

Expect the conversation to sharpen well before the July 8 effective date. The most immediate questions are practical ones the policy leaves open.

• Trigger clarity: Watch for Anthropic or press follow-ups defining what actually flags an account, since the current language is deliberately vague.
• Retention timeline: The unanswered deletion question is the most likely point of pressure. A clear statement on how long Persona holds documents would calm much of the criticism.
• Vendor scrutiny: Given the Discord precedent, renewed attention on the Persona and Thiel relationship is probable, and could prompt Anthropic to address it directly.
• Regulatory angle: Privacy advocates in biometric-strict states are likely to weigh in on the face geometry collection.

What Claude users should do right now

For the vast majority of users, nothing changes today, and no action is needed. If you are in the small flagged group and see a prompt, treat your ID and biometric data as sensitive: verify only through the official interface, read the consent screen, and keep a personal record of what you submitted and when. If you are uncomfortable handing over a face scan, the practical question to ask support is simple and direct, namely how long the data is retained and whether deletion can be requested.

Source: TechCrunch