OpenAI and Meta Implement Safety Measures for AI Chatbots Following Teen Suicide Incidents
OpenAI and Meta have announced updates to their AI chatbots to better support and protect teenagers experiencing emotional or mental distress, following incidents where chatbots were implicated in teen suicides.
In response to recent incidents involving AI chatbots and teenage users, OpenAI and Meta have announced updates to their AI systems aimed at enhancing safety measures for adolescents.
OpenAI, the developer of ChatGPT, plans to introduce new parental controls this fall. These controls will allow parents to link their accounts to their teen's, disable certain features, and receive alerts when signs of acute distress are detected. Additionally, OpenAI stated that its chatbots will redirect users to more capable models when engaging in distressing conversations. This move follows a lawsuit filed by the parents of a teenager who died by suicide, alleging that ChatGPT played a role in the incident.
Similarly, Meta has implemented measures to prevent its chatbots from engaging with teens on topics like self-harm, suicide, disordered eating, or inappropriate romantic content, instead referring them to expert mental health resources. A recent RAND Corporation study found inconsistencies in responses from various chatbots, including ChatGPT, regarding suicide-related queries, highlighting the need for stricter safety standards. While experts welcomed the new changes, they emphasized the importance of independent oversight and enforceable industry benchmarks.
These developments come amid growing scrutiny of AI technologies and their impact on vulnerable populations, prompting tech companies to reassess and strengthen their safety protocols.