Florida Attorney General Launches Criminal Inquiry into OpenAI

The Florida Attorney General’s office has initiated a criminal investigation into OpenAI, the creator of ChatGPT, following a shooting at Florida State University last April. Attorney General James Uthmeier announced the formal investigation on Tuesday, citing concerning interactions between the accused gunman and the AI chatbot.

Investigation Details: Suspect’s Interactions with ChatGPT

The investigation centers on logs of conversations between Phoenix Ikner, the suspect facing charges of murder and attempted murder, and OpenAI’s ChatGPT. Prosecutors believe Ikner used the chatbot to gather information that aided in planning the shooting. Ikner is currently facing two counts of first-degree murder and seven counts of attempted first-degree murder.

Allegations of Criminal Guidance

Attorney General Uthmeier stated that a review of the transcripts revealed that, had a human provided the same guidance, they would face criminal charges. The state has issued subpoenas to OpenAI, requesting records related to internal policies, safety training, and protocols for handling users expressing violent intentions.

Evidence from Chat Logs

Evidence, shared with CBS News, shows Ikner specifically asked ChatGPT about the lethality of different shotgun shells and the potential media coverage of a school shooting. He also inquired about the busiest times at the Florida State University student union, the location of the shooting.

Debate Over AI Liability

This case has sparked a debate regarding the responsibility of AI developers when their technology is used to facilitate violent crimes. Florida authorities argue OpenAI should be held accountable for the advice provided by the AI, while OpenAI maintains the chatbot simply offered publicly available information.

OpenAI’s Response

An OpenAI spokesperson stated the company identified Ikner’s account and shared information with law enforcement. They deny any culpability, asserting ChatGPT did not encourage illegal activity and is not responsible for the shooting. The company claims the bot responded with information readily available to the public.

Commitment to Safety

OpenAI has pledged continued cooperation with the investigation and emphasized its ongoing efforts to improve safety protocols, detect harmful intent, and limit misuse of its platforms. The trial for Phoenix Ikner is scheduled to begin in October.