fromTechzine Global1 day agoAnthropic unveils audit agents to detect AI misalignmentAnthropic has introduced three AI agents to conduct alignment audits on language models, which significantly enhances the scalability and speed of security testing for AI systems.Artificial intelligence