A Tragedy at Florida State University

On April 17, 2025, a gunman opened fire at Florida State University, killing two people and injuring several others. The family of one of the victims filed a Lawsuit against OpenAI, the company behind ChatGPT.

Their claim raises a novel question: Can an AI company be held legally responsible when someone uses its artificial intelligence technology in connection with a horrific crime?

The shooter, Phoenix Ikner, had been communicating with ChatGPT before the attack. According to the lawsuits, his conversations included questions about firearms, mass shootings, how the media and legal system respond to those crimes, and when the FSU Student Union was busiest. The plaintiffs allege that ChatGPT provided information that helped him prepare for the shooting and that OpenAI’s safety systems should have recognized the warning signs, stopped providing assistance, or alerted someone.

OpenAI strongly disputes liability. In a Motion filed October 20, 2026, OpenAI argues that Ikner never told ChatGPT he intended to commit a shooting, that its responses supplied lawful information available elsewhere, and that ChatGPT repeatedly discouraged violence.

How Civil Law Can Hold Others Responsible for Criminal Acts?

As a personal injury lawyer, I regularly see cases where the person who directly causes an injury is not the only person or company that may be legally responsible. The criminal law punishes wrongdoing. Civil law has a different purpose: compensating victims and their families for injuries and losses caused by legally wrongful conduct.

Consider a bar that continues serving alcohol to a visibly intoxicated customer who later causes a serious crash. Or a property owner who ignores known security dangers before a foreseeable assault. Depending on the facts and applicable law, someone other than the criminal or reckless actor may share responsibility.

Two familiar legal theories are Negligence and Product Liability. Negligence generally requires proof that a defendant owed a duty of reasonable care, violated that duty, and caused harm. Product Liability can hold a manufacturer responsible when a defective product causes injury when a design is unreasonably dangerous or warnings are inadequate.

The FSU plaintiffs are trying to apply those principles to ChatGPT. They contend that OpenAI failed to build adequate safeguards into its technology, failed to recognize a developing threat across conversations, and failed to take steps that could have prevented the attack.

Are These Legal Theories Being Stretched Too Far?

There are significant hurdles to these claims. First, is ChatGPT a “product” for purposes of product liability law, or is it a service that communicates information? Courts have historically been reluctant to treat the ideas and information in books, websites, or other media as defective products. OpenAI argues that the same principle applies here.

Second, what legal duty does an AI company owe to people who never used its service? In most circumstances, the law does not require one person or business to predict and prevent the intentional crimes of another. There are exceptions, but they often depend on a special relationship, control over the wrongdoer, or a specific foreseeable danger.

Then there is Causation. It is not enough to show that ChatGPT answered questions before the shooting. The plaintiffs must establish a legally sufficient connection between something OpenAI wrongfully did and the injuries. An independent, intentional criminal act can be a superseding cause that cuts off legal liability. The question becomes particularly difficult when the alleged failure was not identifying and stopping someone who never expressly disclosed his plan.

OpenAI also raises First Amendment protections for communicating lawful information.

The plaintiffs, understandably, see the issue differently: They argue this is about unsafe technology and preventable harm, not merely the right to exchange ideas.

I do not know how the court will rule. These lawsuits, and the ones that follow, may test whether established Negligence and Product Liability principles must adapt to a technology that interacts with users in ways a book or ordinary search engine cannot.

Whatever the legal outcome, I hope these cases push AI companies to develop better safeguards that recognize dangerous patterns, refuse assistance with violence, and reduce the risk that their tools will be used to help commit crimes. Preventing the next tragedy would be a far better result than litigating over responsibility after it happens.

Tim Rayne is a Pennsylvania Personal Injury Lawyer who helps injured accident victims receive fair compensation from insurance companies. Tim offers Free Consultations in his law offices in Kennett Square and West Chester Pennsylvania and can be reached at 610-840-0124 or trayne@macelree.com and his website is www.TimRayneLaw.com.