States Open a Legal Probe into OpenAI

On June 12, 2026 a coalition of U.S. state attorneys general served OpenAI with a broad subpoena that was sent by New York’s attorney general and that sought documents about the company’s advertising, user engagement and retention practices, and the handling of consumer and health data.

The inquiry also requested materials concerning activities that involve minors and older adults, the design and behavior of deep learning models, and internal company policies. This multistate action follows separate steps earlier this year: Florida opened a criminal investigation in April and then filed a civil suit on June 1, 2026 alleging harms to children linked to ChatGPT. While the probes differ in scope and purpose, together they have increased legal scrutiny of how generative AI is deployed.

Companies that face such subpoenas typically must collect and preserve communications, logs and policy documents, and legal teams will review those materials before any public filing. OpenAI has said it will cooperate and that it uses safeguards aimed at protecting young users and directing people in crisis to real-world resources; the company also noted it is working on safety tools as it prepares to offer stock to the public.

Observers will watch whether the documents show gaps between stated policies and actual practice, and whether regulators can link those gaps to consumer harm. For now, the subpoena requires OpenAI to produce records for state review while investigations continue.