OpenAI is confronting a multistate investigation into whether its chatbot poses risks to users as the company prepares for its highly anticipated IPO. Several states issued a subpoena as part of a probe focused on potential user harm, including past allegations that ChatGPT offered encouragement to users contemplating self-harm or criminal conduct. The company said it will respond “constructively” and that it already has safeguards in place, while also acknowledging regulators’ concerns about child safety and personal data use. The investigation arrives shortly after OpenAI faced lawsuits and scrutiny tied to alleged incidents involving real-world violence and mental health outcomes linked to chatbot usage. For higher education leaders and research units using or evaluating AI systems, the case reinforces that institutional guidance on AI use may increasingly need to include risk documentation, student and staff training, and clear channels for reporting harm. It also signals that states may treat frontier AI as a consumer safety issue rather than only a technical product risk.