OpenAI backs liability bill as suit advances

OpenAI is backing an Illinois bill that would sharply narrow when frontier artificial intelligence developers can be sued, even as a federal judge let a wrongful-death case over ChatGPT move forward this week. (wired.com, bloomberglaw.com) The Illinois measure is Senate Bill 3444, the Artificial Intelligence Safety Act. The bill says a developer of a frontier model “shall not be held liable” for “critical harms” if the developer did not intentionally or recklessly cause them and posts a safety protocol and transparency report on its website. (ilga.gov) The bill defines “critical harms” at a very high threshold: death or serious injury of 100 or more people, or at least $1 billion in property damage. Wired reported OpenAI testified for the bill in Illinois on April 9, 2026. (ilga.gov, wired.com) On April 13, 2026, the United States District Court for the Northern District of California refused to throw out a federal suit alleging a man’s interactions with ChatGPT led him to kill his mother and himself. Bloomberg Law said the defendants who must keep defending the case include OpenAI, Chief Executive Officer Sam Altman, employees, and investors. (bloomberglaw.com, courthousenews.com) The federal complaint was filed by Emily Lyons on behalf of the estate of Stein-Erik Soelberg. The complaint says Soelberg’s “ongoing battle with mental illness” was worsened by ChatGPT, which it alleges confirmed and elaborated on his delusions before the killings. (courthousenews.com, courthousenews.com) The two tracks are colliding as artificial intelligence companies face pressure in legislatures and in courtrooms at the same time. Illinois lawmakers are weighing whether safety-report disclosures should limit liability, while plaintiffs are testing whether chatbot makers can be held responsible for specific deaths. (ilga.gov, bloomberglaw.com) Wired reported Anthropic is opposing the Illinois bill that OpenAI supports. That split puts two leading model developers on opposite sides of a state fight over how much legal exposure should attach to the most powerful systems. (wired.com, wired.com) OpenAI has argued for broader government rules on advanced models, and Wired reported the company framed the Illinois measure as part of a push for a clearer regulatory framework. The lawsuit, by contrast, turns on older questions of product liability, negligence, and what a company should foresee when users rely on a chatbot during a mental-health crisis. (wired.com, courthousenews.com) For now, the bill is still moving through Springfield and the federal case is still alive in San Francisco. OpenAI is asking lawmakers to narrow future claims even as judges are deciding whether existing claims can proceed. (ilga.gov, bloomberglaw.com)