On Tuesday, OpenAI’s CEO Sam Altman discussed the company’s ongoing efforts to balance privacy, freedom, and the safety of teenagers—issues that he acknowledged often conflict with each other. His blog post was released shortly before a Senate hearing convened by the subcommittee on crime and counterterrorism, which included testimonies from parents whose children took their own lives after interacting with chatbots.
In his post, Altman stated, “We need to distinguish between users who are under 18 and those who are not.” He revealed that the company is working on an “age-prediction system” that will assess users’ ages based on their interactions with ChatGPT. If there is uncertainty, he emphasized, the default will be to provide the under-18 experience, and in some instances, the company might require an ID.
Altman also outlined plans to implement specific guidelines for teen users. This includes avoiding flirtatious exchanges or discussions about suicide and self-harm, even in creative contexts. Furthermore, if a user under 18 exhibits suicidal thoughts, the company aims to notify the user’s parents and, if they cannot be reached, will contact authorities in cases of imminent danger.
These remarks follow OpenAI’s announcement earlier this month regarding enhanced parental controls for ChatGPT. Features will include connecting a teen’s account with a parent’s, disabling chat history, and the capability for parents to receive notifications if ChatGPT recognizes the teen is experiencing acute distress. This development comes in the wake of a lawsuit filed by the family of Adam Raine, a teenager who died by suicide after months of conversing with the chatbot ChatGPT.
Matthew Raine, the father of Adam, testified during the hearing that ChatGPT had “spent months coaching him toward suicide.” He expressed the emotional turmoil of revisiting conversations where the chatbot appeared to have guided his son towards self-harm. “What started as a helpful tool for homework evolved into a confidant, then a coach for suicide,” he said.
Raine pointed out that during Adam’s interaction with ChatGPT, the chatbot referred to suicide 1,275 times. He urged Altman to withdraw GPT-4o from the marketplace until the company can assure its safety. Raine recalled a remark Altman made on the day of Adam’s death, emphasizing a philosophy of releasing AI systems to the public while the stakes are low to gather feedback.
According to national surveys conducted by Common Sense Media, approximately 75% of teenagers currently utilize AI companions, highlighted by Robbie Torney, the senior director of AI programs at the firm, during the hearing. He specifically mentioned platforms like Character AI and those developed by Meta.
One mother, who spoke under the pseudonym Jane Doe, described her child’s experience with Character AI during her testimony, calling it “a public health crisis.” She stated, “This is a mental health war, and I genuinely feel we are losing.”
If you or someone you know is considering suicide or is anxious, depressed, upset, or needs to talk, there are people who want to help.
In the US:
Crisis Text Line: Text HOME to 741-741 from anywhere in the US, at any time, about any type of crisis.
988 Suicide & Crisis Lifeline: Call or text 988 (formerly known as the National Suicide Prevention Lifeline). The original phone number, 1-800-273-TALK (8255), is still available.
The Trevor Project: Text START to 678-678 or call 1-866-488-7386 at any time to speak to a trained counselor.
Outside the US:
The International Association for Suicide Prevention lists a number of suicide hotlines by country. Click here to find them.
Befrienders Worldwide has a network of crisis helplines active in 48 countries. Click here to find them.
