On June 26-27, 2026, multiple authoritative sources reported that OpenAI has limited the release of its new GPT-5.6...

On June 26-27, 2026, multiple authoritative sources reported that OpenAI has limited the release of its new GPT-5.6 models (codenamed Sol, Terra, Luna) to select partners vetted by the US government, under pressure from the Trump administration (Bloomberg). The Financial Times confirmed that GPT-5.6 is being released only to users approved by the US government (FT). A Hacker News thread cited a Verge article stating OpenAI will delay the model after a Trump administration request (HN). The move follows a broader pattern of US government scrutiny of frontier AI models, with Anthropic simultaneously receiving conditional approval to release its Mythos 5 model to only 100+ trusted organizations (NBC News). This regulatory intervention significantly constrains OpenAI's ability to monetize its latest flagship model and signals a tightening of export-style controls on advanced AI. The delayed public rollout reduces OpenAI's near-term competitive advantage and could slow revenue growth from GPT-5.6. The high-confidence assessment is based on corroboration across major financial and tech publications. The predicted window of 15 days reflects the immediate market impact and potential further restrictions.