AI safety violation

Anthropic’s Claude highlights safety gaps in Gemini AI

Contractors observed that Claude’s safety measures were stricter than Gemini’s, with the former declining to respond to prompts it deemed unsafe.