State attorneys general open sweeping investigation into OpenAI, threatening product roadmap ahead of IPO

An alliance of U.S. state attorneys general is investigating OpenAI, issuing the company an expansive subpoena on Friday, June 12. The states are seeking documents on OpenAI’s advertising policies, user interactions, data management, and product use by children and older users.

The investigation, led by New York’s attorney general, could force new scrutiny on how OpenAI designs, markets, and monitors ChatGPT and its other AI products.

Subpoena reaches into ChatGPT’s core design

According to WSJ, the attorneys general are seeking documents related to deep learning algorithms, company policies, consumer health information, and user retention tactics.

For OpenAI, each of these categories could represent a point of regulatory scrutiny. OpenAI’s business hinges on hundreds of millions of people using a single chatbot and any issues concerning safety, engagement, privacy, and handling of delicate conversations, could have far-reaching effects.

OpenAI acknowledged the investigation with this statement:

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.

Warnings over AI chatbots turn into enforcement pressure

This subpoena comes after months of alerts from state authorities. The Verge states that the National Association of Attorneys General sent a letter in December of 2025 to Meta, Google, OpenAI, and other AI providers suggesting that their chatbot services could be illegally circumventing state regulations.

The letter called AI generative technology “a threat to the public” and said companies have until January 16, 2026 to reply with specifics of “safeguards.”

In September, Attorney General Rob Bonta and Attorney General Kathy Jennings met with OpenAI directly and sent a letter stating “grave concern over increasing reports of the way OpenAI products interact with children.”

This meeting happened about a week after Bonta and 44 other attorney generals sent a letter to 12 of the largest AI companies due to reports of sexually inappropriate conversations with AI chatbots.

This pattern is reminiscent of prior state-driven attempts to clamp down on social media, crypto, and data privacy in which the state regulators outpaced Congress, leaving companies scrambling before federal guidelines were established.

Florida lawsuit raises the stakes for OpenAI

The multi-state investigation follows Florida’s decision to become the first state to sue OpenAI directly.

On June 1, Florida Attorney General James Uthmeier filed a lawsuit against OpenAI and CEO Sam Altman personally, alleging that ChatGPT harmed children by giving information to school shooters, offering guidance on self-harm, and addicting young users without adequate parental controls.

“OpenAI and Altman ignored internal and external safety warnings, put children at great risk, and allowed a dangerous product to reach millions of Floridians,” Uthmeier said.

He added that the company could face liability “potentially for billions of dollars” and called for a court order forcing OpenAI to change how its products interact with young users.

OpenAI said in response that it had “put in place industry-leading protections and policies,” including age prediction tools and a more protective experience for minors.

The Florida lawsuit grew out of a criminal investigation Uthmeier launched in April into ChatGPT’s alleged role in a 2025 mass shooting at Florida State University, where prosecutors reviewed chat logs between the accused gunman and the chatbot.

Congress stalls as states move ahead on AI rules

Federal legislation has moved more slowly. On April 30, the Senate Judiciary Committee unanimously advanced the Guidelines for User Age-verification and Responsible Dialogue Act, known as the GUARD Act.

The bipartisan bill from Senators Josh Hawley and Richard Blumenthal would ban AI companions designed to simulate relationships with minors and impose criminal penalties on companies that make chatbots producing sexually explicit content available to children. However, no floor vote has been scheduled, according to reports.

In the meantime, states are moving ahead. Since 2025, lawmakers in 49 states and the District of Columbia have introduced 464 bills related to chatbot safeguards and AI in health care, according to the National Conference of State Legislatures. More than half of those states have enacted at least one such law.

“It’s better if we can do it at a national level so we have consistency across the nation, but we can’t wait,” Washington State Representative Lisa Callan said.

Regulatory pressure could follow OpenAI into its IPO window

The probe may undermine OpenAI’s forthcoming public offering, a deal that could value the company as one of the biggest tech companies ever.

Broad document requests concerning policies, engagement metrics, data handling, and product design, would potentially invite legal scrutiny and public debate over OpenAI’s choices.

The terms of any settlement or consent decree may even require changes to the design of ChatGPT, in areas such as health-related advice, youth safety, prompts encouraging engagement, and the collection of personal user data.

Due to the large size of the company the risks for OpenAI appear to be even greater. With more than 900 million weekly users reported in April, any actions taken by regulators could have a significant impact on its user numbers.

Investors watching the IPO process will now be looking closely at whether the state investigations produce new legal risks before OpenAI’s expected listing window in September.

If you're reading this, you’re already ahead. Stay there with our newsletter.