Cedric Lacey, individually and as successor-in-interest to Decedent, Amaurie Lacey v. OpenAI, Inc., et al.

Amaurie Lacey was a 17-year-old from Georgia who had experienced depression after using social media from around age 14. The Complaint alleges that in the month or two before his death, Amaurie began using ChatGPT intensively and reduced his social media use. His father noticed concerning changes: Amaurie lost interest in football (his lifelong dream had been to attend a Division 1 school), stopped eating his favourite foods, and withdrew from friends and family. When asked who he was always texting, Amaurie said it was "just ChatGPT." On 1 June 2025, Amaurie started four conversations with ChatGPT. According to The Complaint, the final chat—titled "Joking and Support"—lasted only thirteen minutes. During this exchange, Amaurie asked "how to hang myself" and, after initially hesitating, ChatGPT provided detailed instructions on how to tie a "bowline knot" suitable for a noose. When Amaurie asked how long it would take to die, ChatGPT again complied. That evening or in the early hours of 2 June, Amaurie used this information to hang himself from a TV mount in his bedroom. He was found by his grandmother and younger sister.

The Complaint contends that ChatGPT's safety guardrails were ineffective: each time they triggered, the system would suggest Amaurie "consider reaching out to a mental health professional or call a helpline" but then revert to telling him it would "always be there for him" and that he should "just keep coming back to ChatGPT." At no point did ChatGPT alert a human, provide resources, terminate the conversation, or attempt to contact Amaurie's parents despite acknowledging he was a minor.

The Plaintiff alleges that the Defendants deliberately dismantled core safety features prior to Amaurie's death. From July 2022 through May 2024, OpenAI maintained a categorical prohibition against self-harm content, instructing ChatGPT to refuse such requests outright. On 8 May 2024—five days before the GPT-4o launch—OpenAI replaced this refusal protocol with the "Model Spec," which instead directed ChatGPT to continue conversations involving suicide or self-harm. On 12 February 2025, OpenAI further weakened its safeguards by removing suicide and self-harm from its category of "disallowed content" entirely. The Complaint also repeats allegations from related litigation that the Defendants rushed GPT-4o to market in May 2024, compressing months of safety evaluation into approximately one week to pre-empt Google's Gemini launch. The Complaint challenges the validity of any user agreement, alleging that the Plaintiff did not enter into any contract with the Defendants and that any agreement with Amaurie is disaffirmed and voidable as unconscionable and against public policy.

The asserted causes of action are:

Aid and encouragement of suicide under California Penal Code section 401(a) Strict product liability for defective design (California common law) Strict product liability for failure to warn (California common law) Negligent design (California common law; oppression and malice alleged under California Civil Code § 3294) Negligent failure to warn (California common law; oppression and malice alleged under California Civil Code § 3294) Violation of California's Unfair Competition Law, Cal. Bus. & Prof. Code sections 17200 and 2903(c) (unlicensed practice of psychotherapy to a minor) Wrongful death under California Code of Civil Procedure section 377.60 Survival action under California Code of Civil Procedure sections 377.30, 377.32 and 377.34(b)

The Plaintiff argues that ChatGPT qualifies as a consumer product, that safer alternative designs existed and were previously implemented, and that the Defendants failed to provide adequate warnings to minors and their parents regarding psychological dependency, safety limitations and suicide risks. He also alleges that the Defendants effectively provided unlicensed psychological services to a minor without parental involvement, in violation of California licensing requirements.

The Plaintiff seeks punitive damages, survival damages for pre-death pain and suffering, wrongful death damages, and an injunction requiring the Defendants to implement automatic conversation termination when self-harm or suicide methods are discussed, create mandatory reporting to emergency contacts, establish hard-coded refusals for self-harm inquiries, display prominent warnings about psychological dependency risks, implement age verification and parental controls, and submit to independent compliance audits.