OpenAI will release GPT-5.6 in a phased manner, reaching USG-approved partners first

OpenAI is preparing the arrival of its next avant-garde model GPT-5.6which will take place soon but will not follow a usual launch course, but will instead be made available to a certain group of partners in a phased mannerbased on the instructions of the United States Government.

Until now, when the company led by Sam Altman has launched a new model on the market, it is usually accessible to all users at a general level from the first moment. However, it is changing strategy with its next artificial intelligence (AI) model, which will follow a similar launch process to what Anthropic has carried out with Mythos.

This is because OpenAI plans to launch GPT-5.6 in a phased mannermaking it available initially from a specific group of business customerswhich must be previously approved by the Trump administration itself.

This was detailed by the CEO of the technology company in a meeting with his employees held this week, as The Information has learned according to statements from internal sources, which have clarified that the US Government will approve access on a “customer-by-customer” basis during the preview period, limited by request of the administration itself.

Following this limited testing period, OpenAI hopes to be able to conduct a general release for all GPT-5.6 users, which will take place “a couple of weeks later” predictably, as the aforementioned media has pointed out.

This change in the release plan is due to the Government concerns of the United States for offering anyone access to high-capacity American models, with the idea of keep your most advanced technologies reserved and prevent them from being used for malicious purposes or by foreign groups.

Thus, it must be taken into account that the US president, Donald Trump, recently signed an executive order requesting the AI companies to allow federal review of their most powerful models before they are launched public.

However, Altman himself has shared that he is not satisfied with this type of release. “We have made it clear to the US government that This is not our preferred long-term model.and we will work with them and other industry players to achieve a more sustainable approach for future launches,” he said, according to The Information.

It should be noted that this procedure is similar to that already carried out by Anthropic with its Claude Mythos model, which from the beginning has only been available to partners who are part of the Glasswing Project, because its high AI capabilities can pose a cybersecurity risk in the wrong hands.

Likewise, Anthropic recently had to restrict access to two of its most advanced models, Fable 5 and Mythos 5, due to a directive issued by the United States, due to a access to this technology by a group linked to Chinaconsidering it a risk to American security.