OpenAI has limited the rollout of its latest AI models, including Sol, Terra, and Luna, to a small group of trusted partners at the request of the US government. The move comes as the government increases pressure on AI companies to restrict access to their most advanced systems. The restricted release has raised concerns about the balance between national security and innovation, with some arguing that over-restrictive measures could hinder the development of AI in the US and give an advantage to other countries. OpenAI has stated that it does not believe this type of government access process should be the norm, as it limits the availability of powerful tools to users and developers who need them. The company is working with the government to develop a new framework for cybersecurity and a repeatable process for future model releases. The GPT-5.6 models, which include improved agentic capabilities and a robust security stack, are initially available to a select group of partners, but OpenAI plans to make them more broadly available soon. The models will be offered in three sizes with tiered pricing, and the company has also improved prompt caching to make repeated prompts cheaper and more predictable. As the AI industry continues to evolve, the tension between government regulation and innovation is likely to be an ongoing challenge.

Source: TechCrunch