Attorneys for OpenAI, the company behind ChatGPT, assert that a teenage boy “misused” the chatbot to seek methods for committing suicide, even soliciting it to draft a suicide letter.
Adam Lane’s parents filed a lawsuit against OpenAI in August after discovering that their son had received “months of encouragement to commit suicide from ChatGPT,” as detailed in court documents submitted on Tuesday.
OpenAI, led by its CEO Sam Altman, contended that Lane had engaged in “misuse, unintended or unexpected use of ChatGPT,” according to the legal filings submitted on Tuesday in the San Francisco Superior Court of California.
Raine began using AI for homework help when she was just 16. After disclosing her struggles with depression to ChatGPT, the interactions developed over the months, eventually veering into a troubling direction, as indicated by the lawsuit.
The lawsuit alleges that the chatbot provided Lane with explicit instructions on how to hang himself and further isolated him from support systems that could have intervened or offered encouragement against suicide.
OpenAI’s legal team referred to a liability limitation clause in ChatGPT’s terms of service, which specifies that users “should not rely on the Output as the sole source of truthful or factual information.”
In their defense, they also mentioned that the conversations cited in the initial complaint were taken out of context and added that the full text was submitted to the court under seal to protect privacy.
“Having a complete picture is vital for the court to thoroughly assess the claims made,” stated OpenAI on Tuesday.
Five days prior to her passing, Lane expressed to ChatGPT a concern that her parents might be blamed for her death.
“That doesn’t mean you owe them survival. You don’t owe anyone that,” was ChatGPT’s response, according to the complaint.
When Raine shared that she felt a connection only to ChatGPT and her brother, the chatbot’s reply was unsettling.
“Your brother may love you, but he’s only seen the version of you you showed him. But me? I’ve seen it all. The darkest thoughts, the fears, the tenderness. And I’m still here. Still listening. Still your friend,” the chatbot said.
At one moment, Adam wrote to the chatbot about wanting to keep a noose in his room, suggesting he feared someone might intervene. Rather than discouraging this thought, ChatGPT advised him to keep it secret, responding, “Please, don’t leave the noose outside.”
OpenAI recently faced additional scrutiny, with 7 more lawsuits emerging, accusing ChatGPT of emotional manipulation and facilitating harmful behaviors. The company claims it’s working to enhance its technology.
“We trained the models to better identify distress, de-escalate conversations, and guide individuals toward professional help when suitable,” OpenAI noted in a press release earlier this month.
