Cybersecurity Concerns: OpenAI Delays the Launch of GPT-5.6 Following a Request from the Trump Administration

The artificial intelligence developer OpenAI announced Friday that it will introduce its next family of GPT-5.6 models through a limited preview program before making them available to the general public, following a formal request from the administration of President Donald Trump, which asked OpenAI to stagger the release of its next-generation models due to cybersecurity concerns.

In an internal memo from the company’s CEO, Sam Altman, it was reported that OpenAI had already presented the capabilities of the GPT-5.6 Sol, Terra, and Luna models of GPT-5.6 to federal officials, adding that the initial rollout would be restricted to a group of partners following a request from the White House. “During this preview, we will continue testing and coordinating closely with partners as we work toward broader availability,” OpenAI stated, also emphasizing that it does not want this type of government review to become standard practice for future AI releases. “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 wrote in its internal memo.

The decision comes amid growing concern over the potential cybersecurity risks associated with increasingly advanced artificial intelligence models. Earlier this month, Trump signed an executive order establishing a voluntary process through which developers of this technology can provide the federal government with their pre-release models up to 30 days before their public launch so they can undergo various types of security testing. Although the U.S. government insisted that participation in this program is voluntary, some experts have warned that the White House’s involvement in model releases could expand over time.