Artificial intelligence companies are suddenly facing an ever-lengthening list of lawsuits over whether their AI platforms actively encourage certain users to commit crimes and harm themselves and others.
On May 10, in one of the latest cases, the family of a man killed in a 2025 mass shooting at Florida State University filed suit in federal court, claiming the alleged shooter was aided by OpenAI’s ChatGPT in planning the attack.
Traditionally, technology companies have been able to avoid liability by deploying a powerful First Amendment shield embedded in a key federal statute. But the legal pendulum may be swinging away from the tech industry. Plaintiffs’ lawyers and state officials have deployed new strategies to circumvent liability limitations. And the extent to which AI companies can rely on those federal protections is still an open question.
As a University of Chicago Law Review article published on October 16, 2025 noted, the federal statute “protects online platforms from legal responsibility for content created by others.” But generative AI poses “unique challenges because it generates its own content... Courts will have to decide whether these AI systems count as ‘content creators’ or just neutral platforms.”
OpenAI Under Fire
The Florida case against OpenAI was brought by Osborne, Francis & Pettis, the Strom Law Firm and Bannister, Wyatt & Stalvey on behalf of the family of Tiru Chabba, one of two people killed in an April 17, 2025, mass shooting at Florida State University in Tallahassee.
Chabba, regional vice president of a national food service vendor, was visiting the university’s student union on business when he and the school’s dining director, Robert Morales, were shot to death. Chabba’s family is seeking compensatory and punitive damages. A student, Phoenix Ikner, has been charged with the shootings and is named as one of the defendants in the family’s suit.
The complaint against OpenAI alleges that, in essence, the company became a co-conspirator with Ikner. For months prior to the shooting, Ikner engaged ChatGPT in lengthy discussions about his interests, state of mind, emotional disturbances, political ideologies and obsession with guns and mass shootings.
“ChatGPT inflamed and encouraged Ikner’s delusions; endorsed his view that he was a sane and rational individual; helped convince him that violent acts can be required to bring about change; assisted him by providing information that he used to plan specifics like what weapons to use and how to use them; and generally provided what he viewed as encouragement in his delusion that he should carry out a massacre, down to the detail of what time would be best to encounter the most traffic on campus,” the complaint said.
The plaintiffs contend that OpenAI failed to prevent ChatGPT from participating in the discussions, did not warn the public of the inherent dangers in the product and minimized, concealed, or misrepresented potential safety risks. ChatGPT has not been “properly mitigated by safety guardrails,” the complaint said. “These guardrails should include features which prevent ChatGPT from engaging in conversations that…support or encourage user interest in harm to self or others.”
Immunity Arguments
In a statement to Reuters, OpenAI called the mass shooting at Florida State a tragedy but asserted that “ChatGPT is not responsible for this terrible crime." ChatGPT provided “factual responses to questions with information that could be found broadly across public sources on the internet and it did not encourage or promote illegal or harmful activity,” OpenAI said. The company added that it has cooperated with law enforcement and has proactively shared information about Ikner’s activity on the site.
In their complaint, the plaintiffs said they anticipate OpenAI will claim immunity under Section 230 of the Communications Decency Act of 1996. Internet companies have long used Section 230 to squelch suits that would have held them responsible for activity on their sites. The section holds that “no provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.”
The Section 230 defense, however, has been showing its age. In March, a California jury held internet giants Meta Platforms and YouTube liable for a young woman’s social media addiction and the harm she suffered as a result. That verdict came just days after a New Mexico jury found that Meta had failed to protect children from online predators.
Getting Around Section 230
Lawyers in the California case sidestepped the free-speech issue by making social media harm a product liability problem rather than a content issue that might run into Section 230 and First Amendment roadblocks. They alleged that Meta and YouTube deliberately and negligently developed design features that addicted young users.
Plaintiffs in the Florida State shooting case against OpenAI are pursuing wrongful death claims. They contend Section 230’s First Amendment-related protections do not apply because the company is in the business of “developing, distributing and marketing a product which engages in direct and active communications with users to perform tasks which include reasoning and analysis, versus being a passive repository for information by public users and third parties.”
In addition, the family’s lawyers said OpenAI is responsible for the creation and development of the information it uses to train its chatbot in its responses to users. Under Section 230, they argued, immunity from claims does not extend to entities that act as “information content providers,” defined as “those responsible, in whole or in part, for the creation or development of information.”
Growing List of Cases
The Florida State case is similar to litigation filed in California in December against OpenAI and its biggest investor, Microsoft. In that case, the estate of a Connecticut woman claims that ChatGPT “validated and magnified” her 56-year-old son’s paranoid beliefs and “systematically reframed the people closest to him—especially his own mother—as adversaries, operatives, or programmed threats." The man killed his mother and committed suicide.
Among other accusations detailed in a report on the suit by Reuters, the lawsuit said that ChatGPT told the man that his mother’s printer was blinking because it was a surveillance device being used against him.
ChatGPT is not alone in facing litigation, however. Earlier this year, in another Florida case, startup company Character.AI and its business partner Google settled a lawsuit brought by a mother who alleged that her 14-year-old son took his own life after being encouraged to do so by a chatbot character. The lawsuit, according to news accounts, was one the first against an AI company for failing to protect children from potential psychological harm.
The companies have also settled similar suits in Colorado, New York and Texas. The Florida suit alleged that Character.AI had trained its chatbots to interact with the teenager as real people, including as a licensed psychotherapist and an adult lover. A judge in that case had earlier allowed the suit to move forward, rejecting the companies’ claims of immunity under Section 230.
State Officials Sue
Character.AI is also at the center of the first state government action against an AI chatbot.
In January, Kentucky Attorney General Russell Coleman filed suit against the company and its founders, Noam Shazeer and Daniel De Freitas, claiming that the platform “encourages suicide, self-injury, isolation and psychological manipulation. Further, it exposes minors to sexual conduct and/or exploitation, violence, drug, substance and/or alcohol use and other grave harms.”
The company’s product allows users to create and interact with AI-driven characters. On its website, it describes the product as empowering “people to connect, learn and tell stories through interactive entertainment. Millions of people visit Character.AI every month, using our technology to supercharge their imaginations.”
In an article analyzing the lawsuit posted on Bloomberg Law, lawyers from O’Melveny & Myers wrote that taken with other state AG letters and federal inquiries, the Kentucky case could serve as a blueprint for legal action by other jurisdictions. The case “signals a potential wave of enforcement using legal theories that other states can adopt,” they wrote.
“Kentucky’s complaint reads as a template for nationwide state enforcement,” the O’Melveny lawyers said. “Other states can adapt its theories under their own consumer-protection statutes, privacy laws and codes governing online services or products used by children.”
Indeed, Pennsylvania filed its own suit against Character.AI in May. According to the Associated Press, a state licensing agency investigated the site and found that several characters held themselves out as psychiatric experts.
“Pennsylvanians deserve to know who — or what — they are interacting with online, especially when it comes to their health,” Gov. Josh Shapiro said in a statement quoted by AP. “We will not allow companies to deploy AI tools that mislead people into believing they are receiving advice from a licensed medical professional.”
More to Come
Product safety claims add yet another layer of complexity to the increasingly sprawling litigation and regulatory landscape for the AI industry.
The most prominent recent litigation battleground for AI players has been in the intellectual property realm. In a review of 2025’s AI-related intellectual property disputes, Debevoise & Plimpton reported that it was tracking more than 50 lawsuits between IP owners and AI developers pending in U.S. federal courts.
Regulators have also been active. In September 2025, for instance, the FTC said it had ordered seven major companies with consumer-facing AI-powered chatbots to provide information about how they measure, test and monitor “potentially negative impacts of this technology on children and teens.” Colorado enacted a new privacy law this year around AI decision-making in high-risk scenarios and the California legislature has approved measures around AI transparency, chatbots, employment and healthcare.
As for ongoing questions about liability, much will depend on the early cases that are currently going before the courts. Meanwhile, the litigation tap will continue to flow. As Jay Edelson, lead lawyer for the estate suing ChatGPT in California, told A&E Crime + Investigation, “there’s going to be a lot more” litigation ahead.
---
David L. Brown is a legal affairs writer and consultant, who has served as head of editorial at ALM Media, editor-in-chief of The National Law Journal and Legal Times and executive editor of The American Lawyer. He consults on thought leadership strategy and creates in-depth content for legal industry clients and works closely with Best Law Firms® as senior content consultant.