OpenAI will delay GPT-5.6 after Trump administration request
TL;DR
OpenAI is not launching GPT-5.6 broadly. The rollout starts as a limited preview for a small group of enterprise partners after a Trump administration request. During the preview, customer access is reportedly reviewed case by case by the US government. OpenAI says this kind of access process should not become the long-term default. The lineup includes Sol as the flagship model, Terra for everyday use, and Luna as the cheaper option. OpenAI is pitching gains in coding, biology, and cybersecurity.
Nauti's Take
The safety concern is real, but the process looks like improvised industrial policy. Models at this level need transparent thresholds, auditable testing, and equal rules for OpenAI, Anthropic, and everyone else.
Case-by-case government approval does not cleanly solve the safety problem; it makes the market more political. OpenAI still benefits because it gets to keep moving through a preview instead of being fully stopped.
Briefingshow
This is more than a launch delay. If frontier models effectively need political pre-clearance, power shifts from product and safety processes toward government gatekeeping. Without clear standards, the result can be arbitrary delays, favored vendors, and uncertainty for developers, enterprises, and global partners.
Sources
