TechCrunch · Security·8d agoResearchers used Anthropic’s Claude to hack into OpenAI#ai-security#anthropic#bug-bounty in the wild 4 min1