Florida Attorney General Opens Investigation Into OpenAI Over Security Threats

Florida Attorney General James Uthmeier has announced a formal investigation into OpenAI, the company behind ChatGPT, the world's most popular chatbot. The state's complaints span several areas: threats to national security, the possible leak of advanced technology to US adversaries, and a direct link between the chatbot and dangerous, unlawful behavior.

In an official statement, Uthmeier said there are serious concerns that OpenAI's data and technology are "falling into the hands of America's enemies, in particular the Chinese Communist Party." The rhetoric is familiar from US technology-security politics, from TikTok to Huawei. Even so, the attorney general insists his office has concrete grounds for the inquiry, not just political motives.

Beyond the geopolitical risks, Uthmeier pointed to ChatGPT's alleged connection to unlawful acts. According to him, the service was "complicit in criminal behavior" involving child sexual abuse material, as well as the "encouragement" of self-harm among users. These are far from the first accusations of this kind against large language models. How effective AI safety filters are, and how they respond to dangerous requests, has long been a source of sharp dispute in regulatory and academic circles. OpenAI has repeatedly updated ChatGPT's safety policies, but critics consider those measures insufficient.

The most resonant allegation is that ChatGPT may have "helped" a person suspected of the shooting at Florida State University in April 2025. The details of the episode have not yet been fully disclosed: it is unknown exactly how the chatbot was used, or to what degree interacting with it may have influenced the suspect's actions. Still, the very fact that the attorney general made such a statement gives the investigation considerable political weight and draws wide media attention.

OpenAI has long been under pressure from both federal regulators and individual state authorities. Over the past two years, the company has turned from a nonprofit research organization into a full-fledged commercial business valued at more than $300 billion. That structural change itself prompted several investigations, notably a suit by the California attorney general challenging the legitimacy of the move to a commercial model. The Florida investigation opens yet another legal front in an already complicated regulatory picture.

It is worth stressing that the investigation is still at an early stage. The attorney general's statement contains no specific charges or documented evidence; it is more a declaration of intent. Holding the company actually accountable would take months of evidence gathering and, likely, court hearings. Still, the publicity creates additional reputational risk for OpenAI at a time when the company is actively pursuing government contracts and partnerships with US federal agencies.

The Florida investigation reflects a broader trend: state attorneys general are increasingly going after major AI companies without waiting for federal regulation. With no single AI law at the US level, the states are becoming the first line of oversight. For the industry as a whole, this means a fragmented regulatory environment in which each state can set its own requirements and interpretations, and where a single chatbot can be the subject of investigations in different parts of the country at once.