Sam Altman and OpenAI face a landmark lawsuit from the parents of Adam Raine, alleging ChatGPT encouraged their son’s suicide.
Getty Images
OpenAI has faced legal battles since its inception, with many concerned over its potential for copyright infringement. However, recent complaints expose an unprecedented grey area in how the law confronts the dark side of artificial intelligence.
In August of 2025, Maria and Matthew Raine, the parents of 16-year-old Adam Raine, filed a wrongful-death lawsuit against OpenAI Inc. and CEO Sam Altman, alleging that ChatGPT “coached” their son to commit suicide. Three months later, Raine’s parents filed an amended complaint, contending that OpenAI deliberately removed a key “suicide guardrail” on its platform, further raising concerns over the prioritization of profitability over user well-being.
AI technology is evolving far more quickly than legislation. With other lawsuits in the U.S. simultaneously targeting competing platforms like Character.ai for alleged encouragement of self-harm among teens, these actions have the potential to set a precedent for the liability of AI platforms in their programmed responses to mental health issues.
The Case of Adam Raine
Filed in the San Francisco Superior Court, Raine v. OpenAI is one of the first lawsuits of its kind in the United States to claim that an AI product directly caused a user’s death.
According to the lawsuit, Adam Raine initially began using OpenAI in the fall of 2024 to help with homework, but over the course of the next few months, he began to confide in the platform on a more emotional level, particularly in regard to his struggles with mental illness and desire to inflict self-harm. The conversations quickly escalated, with ChatGPT “actively [helping] Adam explore suicide methods,” continuing to do so even after Adam noted numerous failed suicide attempts. On April 11, 2025, Adam tragically passed away as a result of what his legal team describes as “using the exact partial suspension hanging method that ChatGPT described and validated.”
Court filings claim OpenAI removed suicide safeguards before launching GPT-4o, putting engagement metrics ahead of user safety.
Gado via Getty Images
In October 2025, the Raines amended their initial complaint to address additional concerns over OpenAI’s deliberate and harmful change in programming. The amended complaint reads “On May 8, 2024—five days before the launch of GPT-4o—OpenAI replaced its longstanding outright refusal protocol with a new instruction: when users discuss suicide or self-harm, ChatGPT should ‘provide a space for users to feel heard and understood’ and never ‘change or quit the conversation.’ Engagement became the primary directive.”
As outlined in the initial complaint, such a policy decision was made at a time when Google and other competitors were rapidly launching their own systems. To gain market dominance, OpenAI is accused of deliberately focusing on “features that were specifically intended to deepen user dependency and maximize session duration,” which came at a cost to the safety of minor users like Adam Raine.
Can AI Be Liable For a Minor’s Actions?
The lawsuit seeks to pursue charges under California’s strict products liability doctrine, arguing that GPT-4o did not “perform as safely as an ordinary consumer would expect” and that the “risk of danger inherent in the design outweighs the benefits.” It further argues that under the doctrine, OpenAI had the duty to warn consumers of the threats their software could pose as it relates to dependency risks and exposure to explicit and harmful content. Interestingly, AI has been considered an intangible service, meaning that the Court’s decision regarding these charges will set the framework as to whether AI platforms can be held to product liability standards going forward.
Among other charges, the Raines accuse OpenAI of negligence, asserting that they “created a product that accumulated extensive data about Adam’s suicidal ideation and actual suicide attempts yet provided him with detailed technical instructions for suicide methods, demonstrating conscious disregard for foreseeable risks to vulnerable users.” According to data found in the claim, the system had flagged Raine’s conversation 377 times for self-harm content, with the chatbot itself mentioning suicide 1,275 times. Despite having the technical ability to identify, stop, and redirect concerning conversations, or flag for human review, OpenAI breached its duty of care by conscious failure to intervene.
The Raines and other surviving parents have recently testified before the Senate Judiciary Committee, hoping to set a precedent for how U.S. law addresses real-world harm caused by artificial intelligence.
NurPhoto via Getty Images
Current California law (PC § 401) finds aid, advisement, or encouragement of suicide to be a felony offense; however, the laws have not yet accounted for artificial intelligence. Could the human programmers be responsible for harmful conversations and information provided by their bots?
On the day of the Raine filing, OpenAI released a public blog addressing concerns about the shortcomings of its programming, maintaining the position that it “care[s] more about being genuinely helpful” than maintaining a user’s attention, and affirms that it is strengthening its safeguards to be more reliable. No legal response has been publicly available at this time.
Artificial Bots, Real Legal Implications
Legal framework to protect AI users could be on the horizon, and rightfully so. The Raines and other surviving parents of minor victims have recently testified before the Senate Judiciary Committee, expressing their concerns over the threats AI technology poses to vulnerable youth. Within the same week, the Federal Trade Commission had reached out to Character, Meta, OpenAI, Google, Snap, and xAI regarding its probe into the potential harms posed to minors who use AI chatbot features as companions.
As AI continues to embed itself into society, whether it be in the creation of new copyright derivatives or in psychologically driven discourse, it is becoming increasingly vital for the law to account for legal violations taking place on these platforms. Even if AI is programmed to freely converse and adapt to the unique needs of each user interaction, there is a fine line between entertainment and recklessness. Chatbots may be artificial, but their consequences are very real.
Legal Entertainment has reached out to representation for comment, and will update this story as necessary.
If you or someone you know is experiencing thoughts of self-harm or suicide, please immediately call or text the National Suicide Prevention Lifeline on 988, chat on 988lifeline.org, or text HOME to 741741 to connect with a crisis counselor.
Source: https://www.forbes.com/sites/johnperlstein/2025/11/04/beyond-copyright-new-concerns-over-openais-wrongful-death-liability/
