OpenAI has introduced a new approval requirement for platforms that redistribute access to its models. Under the company's revised Usage Policies, operators of model routers and AI gateways must now obtain prior review and approval from OpenAI before offering third-party access to its models. The updated rules take effect on November 20, 2026, and the policy revision is dated October 21, 2026, leaving affected platforms roughly one month to complete the required process.

What OpenAI Actually Changed

The update targets what OpenAI describes as model-access intermediary platforms—services positioned between developers and the underlying models. Under the revised policy, any setup in which outside parties reach OpenAI's models directly through such a platform now requires explicit sign-off from the company.

The requirement is deliberately scoped. It does not extend to applications that simply use OpenAI models to power their own features. The line OpenAI is drawing separates using the models from redistributing access to them: one category is a customer, while the other functions closer to a reseller. Vetting arrangements of this kind are a familiar structure elsewhere in technology, where platform owners commonly review partners before redistribution is permitted.

Part of a Broader Policy Refresh

The gateway rule did not arrive alone. It is one element of a wider overhaul of OpenAI's Usage Policies. The refresh also clarifies rules on using OpenAI outputs to train other models, and additional updated sections cover protections for minors, restrictions on high-stakes uses, and national-security requirements. The national-security provisions sit alongside the new approval requirements in the October 21 revision.

Because Usage Policies are the terms that govern how OpenAI's models may be used, the revision reaches the platform's developer base as a whole. The approval requirement itself, though, applies only to the slice of that base that redistributes access to others.

The timing tracks with activity in the intermediary space. Multiple AI gateways and routers have recently rolled out multi-provider routing alongside their own safety measures. Multi-provider routing allows a single platform to send requests to different AI models depending on cost, speed, or task.

Why Gateways Became a Policy Question

When a developer connects to OpenAI directly, the company knows who that developer is and can enforce its terms on that account. When access flows through a gateway, the end user is one step removed, and the gateway effectively becomes a middleman with its own customers.

That middleman layer makes enforcement harder for any of the issues covered in the broader refresh. Protections for minors, limits on high-stakes use, and national-security rules all depend on knowing who sits on the other end of a request. An approval process gives OpenAI a checkpoint: before a platform can resell access, the company gets to review how that platform operates.

What This Means for Developers and Intermediaries

The most direct impact lands on the gateway and router operators themselves. Any platform offering direct third-party access to OpenAI models now faces a review it did not face before.

For the majority of developers, the practical change may be small. Companies that build products on top of OpenAI models for their own features fall entirely outside the approval requirement.

The developers to watch are those who reach OpenAI models through a third-party gateway. For them, the near-term question is which side of OpenAI's new line their platform falls on—whether it uses OpenAI models to power its own product or passes third parties through to the models. If their gateway does not secure approval before November 20, 2026, their access path could be disrupted.