US Government Approves OpenAI's Wide Release of GPT-5.6 AI Models

The US administration has lifted restrictions on the launch of OpenAI's new generation of artificial intelligence models, allowing the company to proceed with a broad rollout of GPT-5.6, according to Axios. OpenAI announced that the new series, which includes the models Sol, Terra, and Luna, will be publicly launched on Thursday, with expanded access already available to preview users. While the initial users have not been officially disclosed, it is expected that early access will be limited to paid subscribers.

The GPT-5.6 series consists of three models tailored for different needs. Sol is the most advanced and costly model, designed for complex tasks such as advanced coding, research, professional analysis, and cybersecurity. Terra serves as a mid-tier model for everyday uses including writing, data analysis, and standard coding. Luna is the fastest and most affordable, intended for simpler and quicker tasks.

About two weeks ago, GPT-5.6 was released to a select group of organizations that underwent a vetting process by the US government. Initial access restrictions were imposed due to concerns about the capabilities of advanced AI models, including potential cybersecurity breaches, impacts on critical industries, and the transfer of advanced technologies to countries like China. Following additional testing and meetings between OpenAI representatives and government officials, the company was granted permission for a wider launch.

The evaluation was conducted by the Center for AI Standards and Innovation within the US Department of Commerce. OpenAI sent technical experts to Washington, D.C., to address professional inquiries during the review process.