Anthropic updated its privacy policy earlier this month, including a new clause stipulating that Claude users would be required to verify their age or identity “in certain circumstances” by uploading an image of a government-issued ID, such as a driver’s license or passport, and it didn’t take long for the outrage machine and rumor mill to start churning.
Some people claimed the new age-verification requirement was the harbinger of a dark new era for the internet, one in which anyone hoping to access a particular service must first prove their identity to one of the major tech giants, all of whom would serve effectively as digital bouncers. Such worries were stoked by Anthropic’s tapping of Peter Thiel-backed software company Persona Identities to build out its age-verification mechanisms. Persona sparked a controversy earlier this year after researchers found the company had been checking users’ private biometric data against government watchlists, prompting Discord to scrap its own plan to implement age-verification measures in partnership with the company. Persona’s roster of current partners includes OpenAI, Lyft, Square, Reddit, and LinkedIn.
Meanwhile, others speculated that Anthropic’s new age-verification requirement could enable the company to keep tabs on its users following a recent order from the Trump administration to cut off access to its newest models, Fable 5 and Mythos 5, to all foreign nationals both inside and outside the U.S. The argument there is that if it chose to, Anthropic could appease the government by granting access only to users who are able to prove their U.S. citizenship.
The common thread here is a sense of anxiety that tech companies, under pressure from the whims of the government, might be able to arbitrarily force private citizens to prove their identity before accessing vital online tools or services, in the same way they would before buying a case of beer or boarding a plane. Such a future would obviously be starkly at odds with the free and open ethos upon which the internet was originally built, not to mention with the democratic-utopian ideal frequently touted by many leaders in Silicon Valley, including Anthropic. The company has worked hard to position itself as a leader in safe, responsible, and human-centered AI.
Taking a step back
Anthropic’s latest clash with the Trump administration has very much dominated its media presence lately, making it pretty much inevitable that any new announcements from the company will be interpreted through a political lens. And given its recent ascendency to the very forefront of the AI race, it’s also expected—reasonable, in fact—that its every move will cause some people to worry about their intentions. Having said that, there are some important details about the company’s new privacy policy to keep in mind.
First of all, Anthropic’s plans to implement an age-verification mechanism long predate the Trump administration’s Fable/Mythos ban. While the company’s privacy policy was officially updated on June 08 (and is scheduled to go into effect July 08), it announced back in December that Claude would begin flagging users suspected of being under the age of 18, and that those users’ account would be automatically deactivated.
Four months later, in April, the company said it was working with Persona to build an age-verification mechanism through which some users would be asked to upload an image of government-issued ID, such as a passport or driver’s license, as well as a “live selfie” using a smartphone or other device. Anthropic wrote in that announcement that “Persona is contractually limited in how they can use your data: only to provide and support verification and to improve their ability to prevent fraud. They’re bound to protect it with industry-standard security controls and delete it in line with the retention limits we’ve set and applicable law.” The announcement also added that identity data would not be shared with any third parties or used to train AI models.
Anthropic, for its part, denies any such link between its age-verification policy and Trump’s ban. In response to a disgruntled X post from a user who said they planned to stop using Claude due to Anthropic’s ties with Persona Identities, Thariq Shihipar, a software engineer at Anthropic, said the update “applies only to a small subset of users” and that it was “unrelated to the Fable or Mythos rollout,” and presumably the company’s ongoing deliberations with the Trump administration to lift the ban on those models.
As for the claim that Anthropic could use age-verification as a means of enforcing a ban against foreign nationals from using its models, it’s important to bear in mind that uploading a government-issued ID doesn’t always prove U.S. citizenship: In most states, driver’s licenses don’t include that information, and licenses can also be obtained by foreign nationals staying in the country on a visa. Proving one’s age and proving one’s citizenship, in other words, are not one and the same.
But even if Anthropic’s age-verification mechanisms are entirely unrelated to its ongoing negotiations with the White House, that of course doesn’t make them entirely benign. AI tools aren’t just some obscure corner of the internet; they’re increasingly becoming the very operating system upon which the entire online ecosystem is being built. Any steps that a major AI developer takes to gate access to its tools for any reason should therefore be closely and carefully scrutinized.
