Florida Attorney General James Uthmeier issued subpoenas to OpenAI on Tuesday morning as part of a new criminal investigation into the company's handling of user threats of harm. The subpoenas seek information about how the maker of ChatGPT approaches threats of harm to users themselves and to others.

The criminal investigation will proceed alongside a civil probe that Uthmeier announced on April 9, when he said his office would examine OpenAI and ChatGPT over national security and safety concerns. Uthmeier is investigating whether ChatGPT provided planning assistance to the alleged gunman in the Florida State University mass shooting that left two people dead in April last year.

"We have been looking into the recent FSU shooting, and that shooter's communications with ChatGPT. Our review of that communication has revealed that a criminal investigation is necessary," Uthmeier said.

He described specific interactions he said took place between the chatbot and the alleged shooter. "ChatGPT offered significant advice to the shooter before he committed such heinous crimes. That the chatbot advised the shooter on what type of gun to use, on which ammo went with which gun, on whether or not a gun would be useful in short range," he said.

"If this were a person on the other side of the screen, we would be charging them with murder. We cannot have AI bots that are advising others on how to kill others," Uthmeier added.

The subpoenas request information from March 2024 until this month about OpenAI's policies and internal training materials regarding user threats of harm to themselves and others. They also seek materials from the same time frame on all policies and internal training documents regarding how the company cooperates with and reports crime to law enforcement agencies. The documents request an organizational chart of the company's leaders and senior managers, along with a list of all employees working on ChatGPT.

In early April, an OpenAI spokesperson said, "We build ChatGPT to understand people's intent and respond in a safe and appropriate way, and we continue improving our technology." The company did not immediately reply to a request for comment on Tuesday's developments. More than 900 million people use ChatGPT each week.