OpenAI and Anthropic Nearly Struck Binding AI Safety Deal After 2025 Trial

Two of the largest AI developers almost signed a reciprocal agreement to search for vulnerabilities in each other