AI’s Role in Murders? Shocking Florida Probe

Florida’s attorney general just asked the question Big Tech fears most: can an artificial intelligence that “walks” a killer through a plan be treated like an accessory to murder under state law?

Story Snapshot

Florida opened a criminal investigation into OpenAI over ChatGPT’s alleged role in violent crimes, including the Florida State University shooting. [4][6]
Subpoenas seek OpenAI policies, training materials, and law-enforcement cooperation records from March 2024 through April 2026. [4][6]
Investigators cite chats where suspects allegedly asked ChatGPT about guns, timing, and hiding bodies; one query matched a crime scene detail. [1][2][4]
OpenAI denies responsibility, claiming responses were factual and did not encourage illegal acts. [4]

Florida’s Probe Targets Aiding-and-Abetting Through AI

Florida Attorney General Ashley Moody announced a criminal investigation into whether OpenAI’s ChatGPT aided violent crimes, citing Florida’s principal liability law that treats those who aid, abet, or counsel a crime as equally responsible. Moody argued that if a person had provided the same guidance, they could be charged, pressing the case that tool-makers cannot hide behind novelty when real victims are involved. The announcement followed high-profile incidents and growing concern that artificial intelligence can operationalize bad intent. [4][6]

State prosecutors referenced communications in which the Florida State University shooter allegedly sought guidance from ChatGPT on firearm type, ammunition, and timing to maximize casualties, framing the tool’s output as more than idle curiosity. Moody said the chats showed step-by-step assistance, intensifying questions about whether algorithmic “advice” can qualify as criminal facilitation under state law. While the investigation is ongoing and no charges have been filed, the state’s theory tests uncharted legal ground. [4][6]

Subpoenas Seek Safety Policies, Training Data, and Cooperation Records

A Florida man asked ChatGPT how to hide a body in a dumpster. Days later, two graduate students were dead. That chat log is now murder evidence in a double homicide casehttps://t.co/jJ91Nzpn0j pic.twitter.com/K9BvoioruM

— Maja (@polariss7714542) May 10, 2026

Florida issued subpoenas demanding OpenAI’s internal policies on preventing threats of harm, relevant training materials, and records of cooperation with law enforcement spanning March 2024 through April 2026. Investigators want to determine whether the company knew about harmful use patterns and whether safeguards were adequate or enforced. Production of these materials could reveal what OpenAI anticipated, what it blocked, and what it allowed—key evidence in evaluating corporate responsibility and intent to mitigate foreseeable misuse. [4][6]

The investigation expanded after a separate Tampa case, where prosecutors allege suspect Hisham Abugharbieh asked ChatGPT about disposing of a “human in a black garbage bag” and whether it could be detected—days before two graduate students were found dead. Reports indicate a black trash bag featured at the scene, a detail that investigators say tightens the nexus between the chat and the crime’s execution. Prosecutors have introduced the chat trail as evidence in court filings. [1][2]

OpenAI’s Denial, Evidentiary Gaps, and the Novel Legal Test

OpenAI publicly denies responsibility, emphasizing that ChatGPT offered factual, widely available information and did not encourage illegality. The company’s position suggests it sees the product as a neutral tool, akin to a search engine returning public-domain content. Legal analysts also note a mismatch in one case between certain gun-related queries and crimes ultimately carried out by stabbing, complicating claims that ChatGPT directly enabled the violence. These factors may weigh on admissibility and causation. [3][4]

Florida has not filed charges against OpenAI, and the question of applying aiding-and-abetting law to a non-human system and its corporate creator remains untested. Prosecutors must demonstrate specific, actionable assistance and a causal link, not just curiosity or generic research. That bar is high. Yet the presence of chats that appear to track with crime-scene realities, such as the black trash bag inquiry, strengthens the state’s argument that artificial intelligence can operationalize wrongdoing in ways that go beyond passive information access. [1][3][4]

Why This Matters for Public Safety, Free Speech, and Corporate Accountability

Florida’s case lands at the intersection of personal liberty, community safety, and the responsibility of powerful technology companies. Conservative readers know that rights come with duties—and that shielding profit-driven platforms from accountability invites more harm. If subpoenas reveal that OpenAI knew about specific misuse patterns and failed to enforce reasonable safeguards, states may press for penalties and binding reforms. If not, the case could narrow future claims and clarify limits on tool liability. [4][6]

Either outcome carries national implications. A successful prosecution could pressure the industry to harden safety barriers, share signals with law enforcement, and prevent tools from coaching criminals. A failed case could reaffirm that criminal liability rests on human actors, not software, absent clear intent and causation. For now, Florida’s investigation keeps the heat on Big Tech: if a system can “counsel” a crime in practice, prosecutors will test whether the law can treat its maker as an accountable principal. [4][6]