[Digital Today reporter Yoonseo Lee] OpenAI is delaying a full release of its next flagship model, GPT-5.6, and will begin a limited release for only some corporate customers.

The Verge reported on Wednesday that the Trump administration asked OpenAI to roll out GPT-5.6 in stages over concerns about potential security issues. Sam Altman (샘 알트먼), OpenAI's chief executive, was reported to have told employees that GPT-5.6 would be made available as a limited preview.

The core of the move is how access will be controlled. During the pilot release period, the Trump administration is expected to approve access on a case-by-case basis for each customer, according to reports. OpenAI is narrowing the scope and eligible customers in line with the federal government's request.

The move is seen as more relaxed than a demand issued to Anthropic earlier this month. The Trump administration previously sent Anthropic a final ultimatum to halt access to "Mythos5" and "Fable5". At the time, the administration also issued export control guidelines that included a ban on foreign nationals accessing the technology. The standard also applied to Anthropic employees who are not U.S. citizens.

The issue is that the administration's AI policy direction is moving differently from earlier messaging. The Trump administration has said it would take an approach in AI that "speed wins". It also said it would pursue a program to encourage U.S. AI exports. In practice, the government is directly intervening in the deployment of the latest models by major AI companies and splitting access targets by nationality and customer-by-customer screening.

The shift has already led to unease in the industry. In the Anthropic case, the tech industry raised concerns that the administration's response was overly hardline. This time, OpenAI opted for a limited release instead of a full halt, also showing that different standards are being applied by company.

For OpenAI, it amounts to choosing a compromise that does not completely stop releasing the latest model while keeping pilot operations for corporate customers. But because the administration will approve customer access on a case-by-case basis, the pace of future model distribution and the timeline for expanding customers could be heavily influenced by government decisions.

At the same time, concerns about consistency in U.S. AI regulation are again coming into focus. Critics say control standards are not uniform, with the same administration demanding a halt in one case while allowing limited release in another. This case confirmed that concerns raised in and outside the industry have become reality in a way seen as "unfair depending on the company."

Two points will be watched going forward. One is when OpenAI's limited release will move to a general release stage. The other is whether the Trump administration will apply the same approval system to new models from other AI companies. As deployment of the latest AI models becomes tied not only to technological competition but also to federal pre-approval, launch strategies in the U.S. AI industry are expected to be affected for some time.