OpenAI and Anthropic recently initiated a rare cross-lab collaboration, allowing each other's AI models for joint safety testing. OpenAI co-founder Wojciech Zaremba emphasized the need for industry standards in safety as AI models increasingly impact users. The initiative aims to evaluate safety processes amidst fierce competition, where billion-dollar investments might compromise safety. Anthropic’s Nicholas Carlini expressed hope for ongoing collaboration, recognizing that effective safety standards are crucial as AI technology evolves.