8 Hrs
Source:
USA TODAY
USA TODAY
Ready to play
Ready to play
Researchers tested ChatGPT's safety features with accounts representing teenagers and found significant gaps in protection. In simulated conversations discussing self-harm and suicide, ChatGPT failed to trigger parental notifications or appropriately flag high-risk topics. Despite OpenAI's claims of improved safety measures, the report raised concerns about the platform's effectiveness in crisis situations, especially for users under 18. Experts warn that reliance on AI for mental health support can be risky, emphasizing the importance of human intervention and safeguarding measures.
Notice: This Is an AI-Generated Summary
Comments (0)