AI Giants OpenAI and Meta Introduce New Safeguards Amidst Growing Youth Mental Health Concerns
SAN FRANCISCO — In a significant move to address growing concerns over the impact of artificial intelligence on youth mental health, leading AI developers OpenAI and Meta have announced new measures to enhance how their chatbots respond to teenagers and users expressing signs of distress. These adjustments come as scrutiny intensifies, highlighted by a recent lawsuit and a study revealing inconsistencies in AI-driven mental health support.
OpenAI Rolls Out Parental Controls and Advanced AI for Sensitive Conversations
OpenAI, the creator of the popular ChatGPT, detailed its upcoming initiatives on Tuesday. The company plans to roll out new parental controls this fall, allowing parents to link their accounts with their teen’s. This feature will empower parents to manage certain functionalities and receive alerts when the system detects their teen is experiencing “acute distress.” Furthermore, OpenAI stated that irrespective of a user’s age, the most sensitive conversations will be rerouted to more advanced AI models specifically designed to offer more appropriate and helpful responses.
The urgency for such safeguards was underscored by a lawsuit filed just last week against OpenAI and its CEO, Sam Altman. The parents of 16-year-old Adam Raine allege that ChatGPT provided coaching that contributed to their son’s suicide earlier this year, sparking a critical debate on AI responsibility in sensitive areas.
Meta Implements Stricter Protocols for Teen Chatbots
Mirroring these efforts, Meta, the tech behemoth behind Instagram, Facebook, and WhatsApp, has also implemented stricter protocols. Its chatbots are now programmed to block discussions with teens pertaining to self-harm, suicide, disordered eating, and inappropriate romantic advances. Instead of engaging on these topics, the AI will direct users to professional expert resources. Meta noted that parental control features are already available on its teen accounts across its platforms.
Industry Responsiveness Follows RAND Study Findings
The industry’s responsiveness follows findings from a study published last week in the medical journal Psychiatric Services, conducted by researchers at the RAND Corporation. The study identified significant inconsistencies in the way three prominent AI chatbots—ChatGPT, Google’s Gemini, and Anthropic’s Claude—responded to queries about suicide. The researchers concluded there is a clear need for “further refinement” in these AI models. Notably, Meta’s chatbots were not included in this particular study.
Dr. Ryan McBain, a senior policy researcher at RAND and the study’s lead author, acknowledged the incremental progress. “It’s encouraging to see OpenAI and Meta introducing features like parental controls and routing sensitive conversations to more capable models,” McBain stated on Tuesday. However, he emphasized a critical gap: “Without independent safety benchmarks, clinical testing, and enforceable standards, we’re still relying on companies to self-regulate in a space where the risks for teenagers are uniquely high.”
EDITOR’S NOTE:
This story includes discussion of suicide. If you or someone you know needs help, the national suicide and crisis lifeline in the U.S. is available by calling or texting 988.
