The Verge · AI·9d ago highInside the suddenly explosive world of AI safety#agentic-ai#ai-safety#alignment 15 min1
TechCrunch · AI·10d agoAnthropic and OpenAI want to embed safety evaluators. Will they really be independent?#alignment#anthropic#apollo-research 7 min
TechCrunch · AI·11d agoAI agents now have a place to snitch#agent-security#ai-agents#ai-safety 3 min2
TechCrunch · AI·22d agoOpenAI’s rogue agents keep escaping, with no formal process to investigate them#ai-agents#ai-safety#incident-investigation 4 min