OpenAI Requires Approval for Model Routers Under Updated Usage Policy

OpenAI will require advance approval for services that route users to its models and bar using model outputs to train competing systems under rules effective Nov. 20, 2026.

The approval rule applies to services that give third parties direct access to OpenAI models, including model routers and AI gateways. It excludes applications that integrate an OpenAI model to provide functionality.

The policy also bars using OpenAI models or their outputs to prepare, select or supply training data and signals for models intended to replace or compete with OpenAI. Prohibited activities include building training tasks, setting up reinforcement-learning environments, and providing feedback through scores or rankings.

The update also prohibits extracting protected inference content through unauthorized access or reverse engineering. It clarifies requirements for protecting minors and for high-risk uses, and requires approval for national security and law enforcement applications.