Can You Sue AI for Aiding a Mass Shooting? Florida Is About to Find Out

On April 17, 2025, a gunman opened fire on the Florida State University (FSU) campus in Tallahassee, killing two people and wounding several others. According to a wrongful-death lawsuit filed on behalf of victim Tiru Chabba's family, the shooter, then-FSU student Phoenix Ikner, spent weeks conversing with ChatGPT before the attack.

Court filings allege the chatbot told Ikner what weapon to use, when the student union would be most crowded, and whether targeting a campus with young people would generate more "national exposure" than another target.

In April 2026, the Office of Statewide Prosecution, under Attorney General James Uthmeier, opened a criminal investigation into ChatGPT’s parent company OpenAI based on Ikner's chat logs, then filed a separate civil enforcement suit against OpenAI and CEO Sam Altman in June under the state's consumer-protection law.

"If ChatGPT were a person," Uthmeier said in announcing the probe,"it would be facing charges for murder."

The Chabba family also filed its own wrongful-death suit against OpenAI, naming Ikner as a co-defendant. The criminal probe, the AG's civil suit, and the family's private tort claim are proceeding on three distinct tracks with different plaintiffs and different remedies at stake.

Tying them together is one legal question with no clear answer: does Section 230 of the Communications Decency Act, a 1996 statute written to protect message boards and early internet platforms from liability for what their users post, protect a company whose product generates the harmful content itself?

The Jurisdictional Fight

Before any Florida court can answer that question, it must decide where the case belongs. Uthmeier's office filed an 83-page, 10-count complaint in Highlands County Circuit Court on June 1 pleading consumer protection, public nuisance, and product liability claims under the Florida Deceptive and Unfair Trade Practices Act (FDUTPA). OpenAI's attorneys removed the case to federal court on July 2, arguing that because the complaint cites the federal Children's Online Privacy Protection Act (COPPA) as evidence supporting one of its state-law counts, the case necessarily raises a federal question.

In their notice of removal, OpenAI's attorneys argued the suit "explicitly alleges claims that arise under federal law."

Florida disputes that framing. The state's motion to remand argues it references COPPA violations as context for its FDUTPA claim, not as an independent federal cause of action, and states Florida "expressly disclaims any cause of action under federal law." The distinction isn't academic: keeping the case in Highlands County means a state jury applying only Florida law, but federal court means different procedural rules and, in this case, a notable judge. The case was randomly assigned to U.S. District Judge Aileen Cannon in the Fort Pierce division, who drew national attention for her handling of the classified-documents case against President Trump.

The timing of her assignment has drawn scrutiny. OpenAI's removal filing landed the same month multiple outlets, citing people familiar with the discussions, reported that OpenAI had proposed giving the federal government an equity stake in the company, part of broader talks between Sam Altman and Trump administration officials about a public-facing AI investment vehicle. Nothing in the record ties that reporting to how Cannon will rule, and no party has suggested the two are connected. But commentators tracking the case have noted the optics of a company negotiating an equity stake with the administration appearing before a judge widely seen as favorable to it, even as they caution that the remand question itself turns on ordinary jurisdictional doctrine, not politics.

Briefing on the remand motion has closed, and Cannon has not yet ruled. Her decision on that threshold question is the next concrete milestone practitioners are watching, since it determines which venue ultimately hosts the Section 230 fight.

The Section 230 Question

Section 230 has long shielded "interactive computer services" from being treated as the "publisher or speaker" of information provided by someone else. The doctrine was built for message boards, comment sections, and social feeds where a platform hosts content but doesn't create it.

Generative AI complicates that framework because a chatbot doesn't just host or organize third-party content, but generates the words at issue itself. If ChatGPT told Ikner when the FSU union would be most crowded, as the Chabba complaint alleges, the argument goes that it isn't OpenAI publishing someone else's statement, but instead its own system authoring the statement, arguably placing it outside Section 230's scope altogether.

That's not just theoretical. In Garcia v. Character Technologies, a case in the U.S. District Court for the Middle District of Florida, a federal judge in 2025 allowed strict liability, negligence, and wrongful death claims against Character.AI to proceed past a motion to dismiss, rejecting the argument that the chatbot's output was fully protected speech. The court ruled that Character.AI functioned as a "product" for purposes of claims arising from defects in the app itself, distinct from the third-party content Section 230 was designed to cover. The doctrine was never resolved by a final ruling, as Character Technologies and Google later settled Garcia along with four related cases in January 2026.

But the framework from that decision is the closest thing practitioners have to a signal of how a Florida federal court views the question.

What Florida Actually Alleges

Florida's complaint is notable for what it isn't: it's not primarily a wrongful-death or personal-injury suit. It leads with FDUTPA alongside public nuisance and product liability theories, seeking civil penalties of up to $10,000 per violation, a court order requiring parental consent before OpenAI collects data from users under 13, and damages that Uthmeier said could run into the billions. At the announcement, he said that OpenAI has "chosen profit over public safety.”

That's a structural choice. Attorneys at Benesch Law have compared it to how states once approached tobacco and, more recently, opioid manufacturers: using consumer-protection and public-nuisance law to regulate an industry through enforcement litigation rather than waiting on AI-specific legislation. Both of those earlier waves started with individual suits and scattered state actions before consolidating into the coordinated, multibillion-dollar settlements that reshaped how those industries operated. Florida's suit fits that broader pattern: state attorneys general increasingly lean on existing unfair-and-deceptive-practices statutes to police AI companies. Part of the appeal: those statutes carry per-violation penalties, don't require proving individualized damages, and are harder for defendants to remove to federal court than claims resting on federal law.

What It Means for Practitioners

This isn't a single-issue case, and attorneys across several practices have reason to track it. Product liability counsel will watch whether courts keep treating chatbot output as a "product" subject to design-defect and failure-to-warn theories, following the path Garcia opened. Technology and AI counsel need the Section 230 outcome, since it will shape how every generative AI company structures its terms of service and moderation practices going forward. Consumer-protection attorneys should watch the FDUTPA theory for its portability to other states with similar statutes. And First Amendment practitioners are watching whether courts extend speech protection to AI-generated text at all.

The stakes extend past Florida. More than 20 lawsuits are now pending against OpenAI nationally over alleged chatbot-linked harms, spanning wrongful-death, negligence, and product-liability claims tied to suicides, overdoses, and violence. In response to the wave of suits, OpenAI said it maintains "a zero-tolerance policy for using our tools to assist in committing violence."

Meanwhile, Bloomberg Businessweek has tracked roughly 40 suits filed against chatbot makers broadly since late 2024, a reminder that Florida's case isn't an outlier, but a bellwether for how courts nationally handle a growing wave of similar claims, and for whether state enforcement will become the primary check on the industry over federal legislation.

Looking Ahead

The next concrete trigger is procedural, not substantive. Judge Cannon's ruling on Florida's motion to remand will decide whether the case proceeds in federal court under her supervision or returns to Highlands County. While that ruling alone won't resolve the Section 230 question, it will determine the forum where that fight ultimately plays out.

Florida's suit sits inside a much larger wave. The Character.AI settlements resolved private claims without resolving the underlying doctrine, and OpenAI now faces a comparable stack of suicide, overdose, and violence-related suits across multiple jurisdictions alongside enforcement actions from other states.

Taken together, these cases are likely to produce the first substantial body of AI-liability case law in the coming year or two. That case law will shape not just whether AI companies can be sued, but on what theory, in which forum, and under whose rules. As one law professor tracking the litigation said, ChatGPT's prominence has put it "in the hot seat as the law tries to wrangle this unusual beast."

For now, the Florida case remains what it's been since Uthmeier opened his investigation: the most concrete test yet of whether a statute written for message boards can hold up against a machine that writes its own words.