GPT-5.6 requires US approval per customer

openai has restricted the rollout of its gpt-5.6 model to a select group of partners after the us government requested a case-by-case review of early access customers. according to reporting by the information, commerce secretary howard lutnick personally intervened to warn openai against proceeding without sign-off from multiple agencies, citing concerns raised after recent instability involving anthropic. the decision stems from the trump administration's executive order targeting ai cybersecurity risks and follows the controversy where anthropic was forced to pull its fable model after launch. altman communicated to staff that while this government gatekeeping is not openai's long-term preference, the company must comply with the current review structure to proceed. this incident suggests that federal agencies are actively moving from voluntary guidelines to direct intervention in the release cycles of frontier ai models.