What happened
A coalition of US state attorneys general, led by New York's AG office, served OpenAI with a formal subpoena on June 12, 2026, opening a broad investigation into the company's practices. The subpoena seeks internal documents covering: advertising and user acquisition/retention practices; handling of consumer data and health data; safety of minors and seniors; deep learning model behaviour and sycophancy; and general company policies. OpenAI confirmed it was served and stated it would engage 'constructively'. The investigation follows a recent Florida AG action and comes as OpenAI prepares for an IPO.
Why it matters
This is a material Tier A enforcement action — a formal multi-state investigative subpoena with real compulsory legal force. It signals that state AGs, wielding consumer protection statutes, are becoming a primary enforcement vector for AI accountability in the US in the absence of comprehensive federal AI legislation. The broad scope (data handling, minors' safety, model sycophancy, advertising) suggests AGs are testing the applicability of existing consumer protection frameworks to AI harms. The IPO timing creates additional pressure on OpenAI to demonstrate compliance.
Action needed
AI consumer-facing platform operators should urgently audit: data retention and health data handling practices, advertising claims, minor/senior safety features, and model behaviour documentation. Review state consumer protection law exposure across all jurisdictions. Preserve documents that may be responsive to subpoena-style inquiries.