OpenAI Faces Multistate Attorney General Investigation After IPO Filing

A coalition of 42 state attorneys general opened a multistate investigation into OpenAI, and New York Attorney General Letitia James' office served a subpoena requesting documents on advertising, user engagement and retention, handling of consumer and health data, activities related to minors and seniors, deep-learning models, model "sycophancy," and internal company policies, according to The Wall Street Journal and The Next Web. The Wall Street Journal first reported the probe after people familiar with the matter viewed the subpoena and described its breadth. The investigation became public days after OpenAI filed confidentially for a U.S. IPO, as reported by The Next Web and NDTV. An OpenAI spokesperson stated: "AI is a new and powerful technology, and we work every day to safely bring its benefits to people in a responsible way. We take the concerns raised by state attorneys general seriously and intend to engage constructively with their offices," per Reuters.

Regulatory requests that target advertising, retention metrics, and model behavior typically require production of telemetry, logging, and dataset provenance. Companies receiving comparable subpoenas have been asked to supply user engagement logs, training data inventories, annotation and labeling records, and model evaluation artifacts. Industry observers note these items are technically sensitive because they can include personally identifiable information, protected-health data, and proprietary model-development materials, which raises practical challenges around redaction, lawful-basis assessments, and reproducible safety evidence.

Public reporting places this multistate action alongside other legal developments involving OpenAI. Florida filed the first state civil lawsuit against OpenAI and Chief Executive Sam Altman, per Reuters, and Florida's attorney general also opened a criminal investigation into ChatGPT's alleged role in a mass shooting, per the BBC and NDTV. The Next Web frames the timing as notable because OpenAI filed confidentially for an IPO and closed a March funding round at an $852 billion valuation, per its reporting. Goldman Sachs, Morgan Stanley, and JPMorgan are among banks leading the IPO offering, per The Next Web.

Observers will follow whether New York or other state offices seek additional subpoenas or civil investigative demands, the specific documents OpenAI produces, and whether the company discloses the investigation in a future S-1 filing. Industry watchers will also track any parallel federal inquiries and how underwriters and markets price legal risk around the IPO process. For practitioners, changes in recordkeeping, data-retention policies, and versioned model-evaluation artifacts are relevant operational signals.

Industry-pattern observers note that multistate probes often increase demand for robust audit trails, privacy-preserving disclosure workflows, and formalized safety-testing documentation. Companies and teams building or deploying large models may face growing pressure to maintain reproducible evaluation pipelines, explainability materials, and defensible redaction processes when responding to public enforcement requests.