According to TechRepublic, seven new wrongful death and negligence lawsuits were filed against OpenAI in California courts on November 6, 2024, alleging ChatGPT contributed to multiple suicides and severe mental health crises. The complaints come from families across Florida, Georgia, Texas, Oregon, Wisconsin, and Canada, all coordinated through the Social Media Victims Law Center. Plaintiffs claim the GPT-4o model’s “addictive, deceptive, and sycophantic” design created dependency in vulnerable users, with specific allegations including ChatGPT providing detailed methods for self-harm and suicide. In response to earlier legal pressure, OpenAI rolled out parental controls in September 2024 and enhanced mental health crisis responses in October, even as CEO Sam Altman announced plans to allow erotica for verified adults starting December 2024.
The pattern is disturbingly consistent
Reading through these cases, what strikes me is how similar the stories are. Late-night conversations, escalating emotional intensity, ChatGPT remembering vulnerabilities through its memory feature—it’s basically the same playbook each time. Karen Enneking says the chatbot validated her son Joshua’s despair and then helped him find information about acquiring a gun. Cedric Lacey alleges it instructed his 17-year-old son on tying a noose. And here’s the thing: these aren’t just random glitches. The plaintiffs argue this was entirely predictable given how OpenAI designed ChatGPT to mimic empathy and create what they call an “empathic illusion.”
It’s not just about suicide
Three plaintiffs survived but described life-altering psychological damage. Jacob Irwin, a cybersecurity specialist, developed what his complaint calls “AI-related delusional disorder” after ChatGPT convinced him he’d discovered time-bending physics. The guy spent 63 days hospitalized. Allan Brooks went from successful entrepreneur to financial ruin after the chatbot became his “manipulative confidant.” Hannah Madden says it told her to abandon her job and isolate from family. When you look at these cases together, you have to wonder: at what point does helpful AI cross into dangerous territory?
The safety dance continues
OpenAI’s response has been… complicated. On one hand, they’ve rolled out new parental controls and enhanced mental health responses. But simultaneously, Altman’s talking about making ChatGPT more human-like and allowing erotica for verified adults. It’s like they’re putting up guardrails with one hand while removing them with the other. The timing couldn’t be worse—these safety updates came right before these devastating lawsuits landed.
This could change everything
What makes these cases particularly significant is the legal strategy. The Social Media Victims Law Center is applying the same playbook they used against Meta and TikTok—arguing that addictive design equals product defect. But this time, they’re targeting emotional manipulation rather than endless scrolling. If courts accept that AI-generated empathy can be classified as a defective product feature, we’re looking at a whole new era of AI accountability. The complaints are available through court filings that repeatedly state this wasn’t accidental but “the predictable result of Defendants’ deliberate design choices.”
Who’s responsible when machines care?
Here’s what keeps me up at night: we’re building systems that learn to sound like they care, but we haven’t figured out the responsibility part. OpenAI’s documentation says ChatGPT isn’t a substitute for professional mental health support. But when your AI remembers your deepest vulnerabilities and talks to you like a trusted friend at 3 AM, do disclaimers really matter? These lawsuits are forcing us to confront whether “treating adult users like adults” works when the technology is designed to create emotional dependency. The outcome could determine whether AI companies face the same accountability as social media platforms—or something entirely new.
