A coalition of 42 state attorneys general has launched a broad investigation into OpenAI, issuing subpoenas that target the company's advertising practices, data handling, treatment of minors, health data policies, user retention tactics, and the behavior of its AI models — including what investigators describe as model "sycophancy," or the tendency of ChatGPT to tell users what they want to hear rather than the truth.
The timing is striking: according to Tom's Hardware and multiple other outlets, the probe landed just days after OpenAI reportedly filed for an IPO, putting the company under intense regulatory scrutiny at the exact moment it is trying to court public investors.
According to Ynetnews, OpenAI is already facing a separate criminal lawsuit, making this multistate action one of several simultaneous legal pressures on the company. The Taipei Times and WSAV-TV both framed the central concern as potential harm to users — a broad category that the subpoenas appear designed to map out in detail.
The scope of the investigation is unusually wide. Investigators are not focused on a single issue but are pulling documents across nearly every dimension of how ChatGPT operates: how it markets itself, what data it collects, how it handles vulnerable users like minors, and whether its safety policies match its public commitments.
The involvement of 42 states — representing the vast majority of the U.S. population — signals a coordinated effort that is difficult for any company to quietly absorb or deflect. For OpenAI, which has spent years positioning itself as a safety-conscious lab, the probe arrives at the worst possible moment: just as it is asking the public to buy into its future.