AI
OpenAI limits GPT-5.6 rollout after government request, pledges against long-term restrictions
OpenAI rolled out its new GPT-5.6 models—Sol, Terra, and Luna—to a restricted set of government-vetted partners, signaling tension between federal oversight and rapid AI deployment.
TechCrunch — AI
OpenAI is releasing its next-generation GPT-5.6 lineup to only a "small group of trusted partners" following a U.S. government request. The three models—flagship Sol, balanced Terra, and cost-effective Luna—represent the company's latest capability tier, but all three face rollout restrictions under Trump administration pressure.
The backdrop: mounting government restrictions#
The move reflects intensifying pressure on AI labs to restrict advanced systems. The administration previously ordered Anthropic to limit access to its Fable 5 model, ultimately forcing a full pulldown. A Trump executive order now asks AI firms to submit frontier models for government review up to 30 days before launch—what critics call a "de facto involuntary licensing regime."
OpenAI pushed back against the arrangement in a Friday blog post: "We don't believe this kind of government access process should become the long-term default. It keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them."
The company framed this preview as temporary, pledging to work with regulators on clearer cybersecurity standards and a repeatable process for future releases.
What Sol, Terra, and Luna do#
Sol showcases improved agentic capabilities in coding, biology, and cybersecurity, with a new "ultra" mode deploying coordinated subagents for complex problems. OpenAI claims Sol outperforms Anthropic's Claude Mythos 5 on coding benchmarks while using a third of the output tokens.
Safety is central to the messaging: Sol layers security hardening directly into its core behavior rather than relying on external filters—a lesson from Fable 5's clumsy content routing. The model prioritizes defensive cybersecurity guidance over offensive exploits.
Pricing (per million tokens):
- Sol: $5 input / $30 output
- Terra: $2.50 input / $15 output
- Luna: $1 input / $6 output
OpenAI has committed to broader availability across ChatGPT, Codex, and API access in the coming weeks, pending further regulatory alignment.

