Anthropic could ask Claude customers to confirm their age and identification by importing their government-issued paperwork, based on a brand new model of the corporate’s privateness coverage.
The AI big says the transfer was to permit customers to attraction having their account flagged for doubtlessly fraudulent exercise somewhat than outright banning them, however comes at a time when Anthropic seeks to placate the Trump administration amid an ongoing standoff over who will get entry to the corporate’s AI instruments.
In response to a brand new part in its latest privacy policy revealed earlier in June and set to take impact on July 8, Anthropic says it should ask for a person to show their age or identification “in sure circumstances,” with out offering particular examples.
Whereas Anthropic has lengthy required customers to be over 18 to make use of Claude, the corporate earlier this 12 months launched age-verification checks to adjust to numerous states and nations that require them. Identification checks had been also announced however weren’t mirrored within the firm’s privateness coverage till extra lately.
When triggered, the coverage would require these customers to add a photograph scan of a government-issued passport or driver’s license. Anthropic says it should additionally gather an individual’s selfie picture or video and the particular person’s digitized model as a face geometry template (which some states, like Illinois, think about to be legally protected biometric information). Anthropic says it should additionally hold a file of the verification end result, equivalent to whether or not the person has reached a sure age.
When reached by TechCrunch, Anthropic spokesperson Michael Aciman shared a hyperlink to an X post from Anthropic’s Thariq Shihipar saying that the change applies solely to a “small subset of customers” whose accounts are flagged however not outright banned. (Anthropic’s spokesperson wouldn’t say what number of customers is a subset, however the firm is believed to have tens of millions of users monthly.)
“[Anthropic’s identity verification policy] was up to date on June 17 as an replace to the appeals course of,” stated Shihipar within the put up. “It’s unrelated to the Fable or Mythos rollout.”
Anthropic said it is allowed to require customers to add a duplicate of their IDs for a variety of causes, equivalent to for requiring customers to confirm themselves for creating and administering their Claude account, and imposing its phrases of service, equivalent to to forestall and examine fraud, abuse, and violations of its phrases, together with illegal or legal conduct, and to research and resolve safety points.
The transfer to maintain nearer tabs on who’s utilizing Anthropic’s AI instruments could also be one manner for the corporate to adjust to a wide range of ongoing authorized challenges, regulatory adjustments, and inbound pressures from the Trump administration.
The tech big stays largely at an deadlock with the White Home, greater than per week after Trump officers successfully pressured Anthropic to tug its newest cybersecurity fashions over allegations that an apparent jailbreak might break the fashions’ guardrails. Different studies have pointed to personality clashes between the company and the Trump administration because the better offender of the breakdown in relations.
This newest conflict comes months after the Division of Protection designated Anthropic a “supply chain risk”, apparently in retaliation for not permitting the federal government to make use of its know-how for mass home surveillance or powering absolutely autonomous weapons.
Anthropic stated it makes use of the San Francisco-based firm Persona as its identity-checking provider and that customers could “see a verification immediate when accessing sure capabilities, as a part of our routine platform integrity checks, or different security and compliance measures.”
Anthropic stated that it decides how lengthy Persona will retain its customers’ identification paperwork, however Anthropic’s spokesperson didn’t instantly say when the info was deleted. (For context, Roblox, one other buyer of Persona, says customers’ photographs are deleted “immediately” after they’re processed, limiting the prospect of the data later being leaked or stolen.)
Persona can nonetheless face U.S. authorities calls for for customers’ info that it shops on its servers.
Persona is backed by Founders Fund, an funding agency based by Trump backer Peter Thiel, who additionally invests in Anthropic. The identity-checking agency has confronted criticism from customers for its hyperlinks to Thiel. Earlier this 12 months, Discord selected Persona for its age-verification checks, then shortly reneged following user backlash for selecting Persona.
If you buy by hyperlinks in our articles, we may earn a small commission. This doesn’t have an effect on our editorial independence.
