OpenAI, Anthropic negotiate landmark deal to stress-test each other’s AI models for safety risks: Report

OpenAI and Anthropic are negotiating a landmark deal to stress-test each other’s AI models for safety risks and unexpected behaviour.