Live Updates > Live Update Details
2026-09-21 23:18:11
[Report: OpenAI and Anthropic Discuss Stress Testing of Each Other's AI Models] The Information, citing a source familiar with the matter, reports that OpenAI and Anthropic have discussed a legally binding agreement to stress test each other's AI models. The source stated that earlier this year, the two companies and their lawyers discussed an agreement to conduct a series of tests on their respective models to identify vulnerabilities or potential dangers. The report states that the proposed agreement would grant both companies access to the application programming interfaces (APIs) of each other's commercially available AI models, excluding those not yet released. The source indicated that both OpenAI and Anthropic have guaranteed that they will not retain each other's data during the testing process.
