OpenAI announced on 27 April 2026 that its services have achieved FedRAMP Moderate authorization, enabling U.S. federal agencies to deploy GPT-4o and GPT-4o mini models in government workflows [source]. The authorization covers ChatGPT Enterprise and the API platform, both hosted on Microsoft Azure Government cloud infrastructure.

FedRAMP Moderate is a standardized security assessment framework for cloud services handling federal data that is not classified but requires protection. The authorization follows a third-party assessment of OpenAI's security controls, including encryption, access management, and incident response procedures.

OpenAI stated that federal customers can now use the models for tasks including document analysis, research assistance, and internal communications. The company emphasized that customer data submitted through the authorized services will not be used to train models, consistent with its enterprise data handling policies.

The announcement did not disclose which federal agencies have already begun using the authorized services or provide details on the timeline of the authorization process. OpenAI noted that the FedRAMP authorization applies specifically to the Azure Government deployment and does not extend to its standard commercial cloud offering.

This marks OpenAI's first formal authorization for U.S. government use at scale. Other AI providers including Anthropic and Google have pursued similar compliance pathways, though public disclosures of their FedRAMP status remain limited. The authorization does not address model behavior, accuracy, or hallucination rates in government contexts—factors that remain relevant for agencies evaluating AI deployment risks.

Federal agencies using the authorized services remain responsible for validating model outputs and ensuring compliance with their own data handling and mission requirements.