a multi-state joint investigative team, composed of attorneys general from several states, has officially launched a formal review of the artificial intelligence company openai. according to informed sources, on friday local time, the new york state attorney general’s office issued a subpoena to openai, demanding the submission of internal documents and data covering multiple dimensions, including operations, product design, and user protection.
the wall street journal reported that the subpoena is extremely broad, encompassing not only advertising strategies, user growth mechanisms, and model behavior tendencies—such as “sycophantic responses”—but also how personal data, particularly sensitive health information, is handled, as well as safety measures designed for minors and elderly users. the exact number of states participating in this joint effort, along with their names, has yet to be disclosed.
openai stated that it is fully cooperating with the investigation. a company spokesperson emphasized that, as an emerging technology with far-reaching societal implications, ai development must balance rapid innovation with responsible governance. the company remains committed to expanding access to its technology in a prudent and transparent manner, taking seriously the core concerns raised by state attorneys general, and pledging to maintain constructive communication and collaboration going forward.
bloomberg further quoted the spokesperson as noting that chatgpt currently incorporates multiple safety mechanisms tailored to vulnerable user groups—for example, providing guided support pathways for teenagers and individuals experiencing psychological crises, encouraging them to connect with real-world professional help resources. however, openai declined to elaborate further on the identity of the investigating authorities or the specific scope of evidence collection.
this investigation comes at a critical juncture when legal pressures on openai continue to mount. just days earlier, the company had won a lawsuit against co-founder elon musk—the court dismissed musk’s allegations of breaching the founding agreement, though his lawyers have announced they will appeal the ruling.
yet the legal challenges facing openai are far from over: from copyright infringement disputes to accusations of algorithmic risks contributing to multiple suicide cases; from the march lawsuit filed by merriam-webster and encyclopædia britannica over training data licensing issues, to last november’s class-action suit brought by families of victims alleging that chatgpt exacerbated mental health crises. the mounting legal scrutiny underscores an increasingly pervasive regulatory environment.
notably, florida attorney general ashley moody took the lead earlier this month by filing a civil lawsuit against openai and its ceo sam altman, marking the first instance of a state-level law enforcement agency directly accusing an ai company of endangering children. the complaint alleges that openai ignored both internal and external warning signs, allowing high-risk products to enter domestic settings on a large scale.
meanwhile, openai is also stepping up efforts to repair its public image. in april this year, altman publicly apologized following the temple terrace shooting incident in canada, acknowledging that after banning the accounts involved, the company failed to promptly coordinate with law enforcement, exposing significant shortcomings in its emergency response mechanisms under extreme circumstances.
as the regulatory storm intensifies, openai continues to press ahead with its capital market strategy—this week, it quietly submitted an ipo application. analysts widely believe that the depth and pace of this multi-state joint investigation, combined with the trajectories of other pending lawsuits, could substantially reshape the company’s valuation framework, the scale of its compliance investments, and its future regulatory adaptation path, potentially establishing new governance standards and accountability benchmarks for the entire generative ai industry.
