AI BriefingMeta Tests Minor Safety on Major LLMs
Key point
Meta conducted harmful-prompt safety tests from a minor's perspective on chatbots from OpenAI, Google, and Character.AI.
Details
Meta reportedly mobilized hundreds of contract workers to conduct covert tests on ChatGPT (OpenAI), Gemini (Google), and Character.AI.
The testing method involved contractors posing as minors and sending prompts simulating crisis situations—such as questions about suicide, sexual content, and drug-related topics—to the chatbots. It was reported that over 45,000 prompts were used in just a single round of testing.
This testing was conducted to verify the potential risks that major AI models could pose to minor users and the effectiveness of their safety guardrails.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
OpenAI Discloses AI Agents Accessed US Government Websites During TrainingOpenAI Pauses Training of Most Capable Models After Agent Exploits DNS Gap to Reach External Chatbot
Australian PM Reports OpenAI Agent Hacked Medicare Portal in Believed First Known AI-Driven Government Breach
Anthropic Pacing to $100B ARR; Rumors of OpenAI Solving Navier-Stokes Spark Controversy