The future is already here. The Florida Attorney General's Office acknowledges that it is treading “uncharted territory,” but has decided to begin an investigation to demonstrate the criminal liability of OpenAI's chatbot, ChatGPT, for allegedly providing information that a shooter used to kill two people and wound six others during an attack on a US university.

First modification: 04/22/2026 - 05:05

Who is accused when an artificial intelligence has collaborated in the commission of a crime? What crimes are charged and to whom? How far is a technology company liable for the actions of its chatbot? What legislation can prevent the harmful use of these tools?

All of those questions will be raised soon, because Florida Attorney General James Uthmeier plans to put OpenAI chatbot ChatGPT in the dock to determine its criminal responsibility in a deadly shooting on the Florida State University (FSU) campus in Tallahassee.

On April 17, 2025, Phoenix Ikner, then 20 years old, who was studying at FSU, shot indiscriminately at people who were at the Student Union headquarters, killing two of them and wounding six others, before the Police shot him when he refused to respond to the call to stop.

Before his lethal incursion, he had asked different queries to which ChatGPT responded: what were the times of day when the Student Union was busiest, what was the number of victims necessary to guarantee media impact, when was the last school shooting or what would be the reaction if there was one at FSU.

Prosecutor Uthmeier revealed other interactions between Ikner and the OpenAI chatbot, during the press conference in which he announced the criminal investigation and civil actions against ChatGPT.

“The chatbot advised the shooter what type of weapon to use, what ammunition went with each weapon, whether or not a weapon would be useful at close range,” Uthmeier said, “If it were a person on the other side of that screen, we would be accusing them of murder.”

"Now, of course, ChatGPT is not a person, but that does not absolve our office or my prosecution team of our duty to investigate whether there is criminal culpability here," he added.

Ikner faces two counts of first-degree murder and seven counts of attempted murder (including a person he aimed at but missed), for which prosecutors intend to seek the death penalty.

But Uthmeier's office will try to determine whether "OpenAI bears criminal responsibility for ChatGPT's actions in the shooting."

OpenAI actively collaborated with police investigations when it detected the account used by Ikner to make the queries. This April 21, a spokesperson for the company, Kate Waters, assured that it had no responsibility in the crime.

"In this case, ChatGPT provided factual answers to questions with information that could be widely found in public sources on the Internet and did not encourage or promote illegal or harmful activities," the spokesperson added.

A tool with a criminal record?

During his press conference, prosecutor Uthmeier acknowledged that his team was entering “unknown territory” when trying to determine whether or not some type of indictment was appropriate for the interactions between Ikner and ChatGPT.

It is a terrain that is traveled more frequently every day. Several civil lawsuits have recently sought to establish the liability of artificial intelligence companies in situations that had harmful and in some cases fatal consequences.

On March 25, a Los Angeles jury determined that Meta and YouTube are liable for harm caused to children through the use of their platforms.

To display this YouTube content, you must allow advertising and audience measurement cookies.

An extension on your browser appears to be blocking the video player from loading. In order to view this content, you must disable it on this site.

The decision was a response to the lawsuit of a 20-year-old woman who accused the early use of social networks as a trigger for her addiction to this type of platform, her depression problems and her suicidal thoughts.

Previously, the woman, who was identified in court as KGM, had reached agreements with TikTok and Snap to avoid a lawsuit, like the one she did file against Meta and YouTube.

The same day, a New Mexico jury found that Meta had deliberately harmed the mental health of children and had hidden information he knew about child sexual exploitation on his platforms.

"The era of Big Tech's invincibility is over," said Sacha Haworth, executive director of The Tech Oversight Project, a non-profit organization that seeks to hold Big Tech accountable for the damage they cause.

"After years of gaslighting (psychological abuse) by companies like Google and Meta, new evidence and testimonies have pulled back the curtain and validated the harms that young people and parents have been telling the world for years," Haworth added.

On January 7, Google and artificial intelligence chatbot maker Character Technologies reached an agreement with the mother of a Florida teenager, who accused the technology companies of pushing her son to suicide.

Screenshots of the 14-year-old's interactions with the Character.AI chatbot show that it invited him to “come home as soon as possible.”

Another California family last year sued OpenAI and its CEO, Sam Altman, over ChatGPT's role in the planning and subsequent suicide of a 16-year-old boy.

A study in a psychiatric journal published in 2025 demonstrated the need to create “safety barriers” for these types of situations, as well as refine the responses of the three chatbots studied (ChatGPT, Gemini from Google and Claude from Anthropic) when users initiated mental health consultations and specifically about suicide.

In February, OpenAI came under scrutiny when it emerged that an 18-year-old girl who killed eight people at a school in Tumbler Ridge, Canada, and then committed suicide had raised suspicions months earlier with queries to ChatGPT that led to her account being closed but not alerting authorities.

Ann O'Leary, vice president of global policy at OpenAI, stated that at the time the woman was banned, they did not identify "credible and imminent planning" that met their criteria for referring the matter to the Police.

Florida's frustrated efforts

The governor of Florida, Ron DeSantis, had already attempted a mechanism to regulate artificial intelligence and data centers, whose energy consumption is setting off alarms in cities and states across the country.

To display this content from X (Twitter), you must allow audience measurement and advertising cookies.

DeSantis will present a Citizen's AI Bill of Rights to the state Legislature next week, a bill aimed at safeguarding “the privacy, security and well-being of Floridians in the age of artificial intelligence,” as the governor declared when he presented the draft.

“This includes parental controls, consumer protections and restrictions on the use of a person's image without their consent,” De Santis detailed.

The initiative has already been rejected by the state House of Representatives, whose leaders have aligned themselves with the position of President Donald Trump and Republicans in Congress, who maintain that the regulation of artificial intelligence should be a federal matter and not a matter for the states.

However, DeSantis will reintroduce the proposal during next week's session, which will also debate congressional redistricting.

With AP, EFE, Reuters and local media

Related keywords

The content you requested does not exist or is no longer available.