U.S. Commerce Department Approves Large-Scale Release of OpenAI's GPT-5.6
Alina Collins
The US Commerce Department has cleared OpenAI's GPT-5.6 for broad public release, ending last month's restriction to government-approved entities only. This signals that Washington is building a case-by-case approval regime for frontier AI models — before formal rules even exist.
What did the approval process look like?
Axios, citing people familiar with the matter, reports OpenAI expects to complete the wide rollout this week. Reuters says it cannot independently verify the report.
The green light came only after additional testing and multiple rounds of talks between OpenAI and government officials.
Testing was handled by the Commerce Department's Center for AI Standards and Innovation — the government body responsible for evaluating AI model safety. OpenAI stationed technical staff in Washington to field questions in real time.
Why was last month's launch restricted?
Last month OpenAI was cleared to release GPT-5.6 only to government-approved entities in batches — no broad public access.
OpenAI said at the time that a phased rollout was not its preferred approach. This means → the company and regulators disagreed on timing; OpenAI wanted a single wide launch.
Similar restrictions were previously imposed on Anthropic's Mythos and Fable models — this is not unique to OpenAI but an industry-wide pattern.
Why are formal standards still missing?
AI companies and the government are currently in a case-by-case negotiation phase — each frontier model requires its own round of talks and sign-off.
The Trump administration's latest AI executive order calls for more specific model-release standards, but those standards have not been finalized.
In plain terms = the rulebook is still being written, but enforcement has already started. Every company launching a new model must queue for approval, with no unified timeline — one step at a time.
What does this mean for the AI industry?
This episode establishes a clear signal: Washington is building a de facto case-by-case approval regime for frontier AI releases.
This means → until a formal standards framework is in place, AI companies' launch timelines will remain subject to a regulatory window — not a ban, but a gate on when and to whom a model ships.
This reflects Washington's posture on AI safety: regulate first, formalize later — rather than waiting for the rules to be written before stepping in.
Content is for reference only, not financial advice.