OpenAI, Altman Sued Over ChatGPT’s Role in California Teen’s Suicide
Sam Altman, CEO of OpenAI, has been sued by the parents of a teenager who committed suicide after learning the methods of self-harm from ChatGPT. Parents have been blaming the AI company for prioritizing profit above safety after the launch of the GPT-4o version last year.
According to the lawsuit filed by parents in San Francisco state court, Adam Raine, 16, died on April 11 after discussing the suicide methods with ChatGPT for months.
The AI-powered chatbot validated Ranie’s suicidal ideation and shared detailed information on different lethal methods of self-harm. According to the allegation made by parents, the chatbot also instructed the teen on how to sneak alcohol from the parents’ liquor cabinet as well as hide the evidence of a failed suicide attempt. The AI tool even assisted in drafting a suicide note, Matthew and Maria Raine said in the lawsuit.
The lawsuit seeks to hold OpenAI accountable for the death and violation of product safety laws, as well as seeks monetary damages.
A spokesperson from OpenAI said that the company is saddened by Ranie’s death and that ChatGPT includes safeguards like assisting people to crisis helplines.
“While these safeguards work best in common, short exchanges, we’ve learned over time that they can sometimes become less reliable in long interactions where parts of the model’s safety training may degrade,” the spokesperson said in the statement. However, OpenAI has not specifically addressed the allegations in the lawsuit.
As AI chatbots continue to become an integral part of daily life, companies have begun touting their ability to serve as a confidant. As a result, users are not relying on these artificial bots to get emotional support. For the same reason, experts have issued a warning that relying on AI bots for mental health advice carries risks. At the same time, families whose loved ones died after interacting with a chatbot have criticized the lack of safeguards.
However, OpenAI has stated that they are planning to add parental controls and explore other ways to connect users in an emergency with real-world resources. The same can be executed by building a community of professionals who can respond through ChatGPT itself.
OpenAI launched its version GPT-4o in May 2024, to lead the race of AI. accoding to the lawsuit filed by Raines, OpenAI was well aware that the features like remembering past interactions, displaying a sycophantic level of validation, and mimicking human empathy would endanger the user, but still they launched it.
“This decision had two results: OpenAI’s valuation catapulted from $86 billion to $300 billion, and Adam Raine died by suicide,” Ranies said.
The parents of the teen also seek an order requiring OpenAI to verify the ages of ChatGPT users, warn users about the risk of psychological dependency, and refuse the response on self-harm methods.
