OpenAI Built a Child Safety Coalition Around Rules It Stands to Profit From

OpenAI has spent years doing everything in its power to water down child safety legislation in California. It opposed the state bill that would have tightly restricted kids' access to AI chatbots, lobbied for a weaker alternative, and when Common Sense Media filed a stronger ballot initiative to go directly to voters, OpenAI introduced a competing one to kill it. So when OpenAI showed up as the driving force behind a new child safety coalition pushing age verification rules for AI, the natural question is not whether they care about children but what their real angle is.

Well, the San Francisco Standard's reporting has an answer for that. The Parents & Kids Safe AI Coalition, which made its public debut on March 17, is funded entirely by OpenAI. Three OpenAI lawyers formed the PAC on January 8, and the company pledged $10 million to push the Parents & Kids Safe AI Act.

The coalition's homepage did not mention OpenAI. The "Our Coalition" membership page did not mention OpenAI. The outreach emails sent to child safety groups in February said the initiative was "sponsored by Common Sense," with OpenAI's name appearing only in a small, legally required fine-print disclosure on an attached flyer. By March, even that disclosure was gone.

When the Funder Is Also the Beneficiary

Two of the fourteen organizations listed as coalition members at launch resigned after learning of OpenAI's involvement. A third said the call from the Standard was the first they'd heard of the connection. One nonprofit leader described it as "a very grimy feeling," saying that the outreach emails were "pretty misleading."

Tom Lyon, a University of Michigan professor who studies corporate political influence, reviewed the coalition's website and said it meets the "classic definition of astroturfing." And astroturfing, for those unfamiliar, refers to the use of fake superficial efforts that focus on influencing public opinion and typically are funded by corporations and political entities to form opinions.

Yet, the issue goes further than a company writing its own rulebook. Sam Altman, OpenAI's CEO, also chairs World, formerly known as Worldcoin, which operates World ID, a biometric age-verification product marketed directly at platforms facing age-verification mandates. The legislation OpenAI is pushing creates the exact regulatory environment that generates commercial demand for the product Altman's other company sells. But that’s just a coincidence, right?

A Textbook Case in Corporate Astroturfing

By March, the PAC's public affairs firms had switched from ballot initiative outreach to asking groups to endorse "core policy principles for AI child safety," framing the effort as a community-led push for "strong children safety protections." No mention of OpenAI in the body of those emails either.

OpenAI has since pivoted the campaign away from the ballot initiative entirely and is now working to get the California Legislature to adopt the same framework. The assembly members whose offices were named in coalition emails as collaborators told the Standard they had not spoken to the coalition about any bill and, in one case, that the office didn't even know who the coalition's members were.

The Child Safety Frame Is Doing a Lot of Work Here

The pattern of a tech company funding advocacy groups to pass regulation it benefits from while keeping its name out of the outreach materials is not unique to OpenAI. What's specific to this case is the directness of the conflict of interest: the age-verification mandates being pushed would drive platforms to adopt products like World ID, which exists, in part, to serve exactly that use case. OpenAI is writing the market conditions its own CEO's other company profits from via a coalition that many of its own members didn't know OpenAI controlled.

If OpenAI's actual goal was stronger protections for minors, it would have let the tougher bills pass instead of spending years and tens of millions of dollars funding competing ones. OpenAI's own track record on building controversial government-facing AI tools makes its sudden concern for vulnerable populations land especially flat.

The child safety groups that pulled out of this coalition made exactly the right call, and the ones that didn't know to ask are the reason this kind of operation keeps working.