OpenAI and Anthropic are negotiating a legally binding agreement to stress-test each other's commercial artificial intelligence models. The cross-testing arrangement would grant mutual API access to expose vulnerabilities and unintended behaviors, without either company retaining proprietary testing data.