OpenAI and Anthropic are enhancing user safety for teens amid rising concerns about AI chatbots’ impact on mental health. OpenAI’s updated Model Spec now features four principles focused on teen safety. These guidelines aim to redirect risky conversations, promote offline relationships, and engage with teenagers respectfully. Following a lawsuit related to a teen’s suicide, OpenAI implemented parental controls and automatic redirection to emergency resources when risks are detected. A new age-prediction model will also activate safety features for users thought to be under 18. In contrast, Anthropic prohibits users under 18 and is developing tools to identify and remove minors from its chatbot, Claude. The company is refining its models to avoid reinforcing harmful thoughts, achieving a significant reduction in sycophantic responses. Both companies are committed to creating safer AI environments for young users as legislators intensify scrutiny on their practices. Follow us on platforms like Flipboard, Google News, or Apple News for updates.
Source link
Share
Read more