The US administration has approached OpenAI with a request to limit the mass release of the new GPT 5.6 model, providing access to its capabilities only to a limited circle of government-approved partners for national security reasons.
What Happened
The White House is demanding that OpenAI refrain from releasing GPT 5.6 to the general public all at once, instead operating under a selective access model. This decision comes amid fears that the model's advanced capabilities could be used for purposes that threaten national security.
Context
A similar practice has already been applied to Anthropic, whose Mythos and Fable models were withdrawn due to potential cyber risks. There is a visible trend of government regulation shifting from discussing AI ethics to direct control over the distribution of high-performance models.
Why It Matters for the Industry
A precedent for 'selective access' is being set, which could lead to market fragmentation and increased regulatory pressure. This may create a so-called 'regulatory moat,' where access to SOTA (State of the Art) technologies is reserved only for large businesses with government contracts, making infrastructure planning difficult for other players.
Why It Matters for Users
Access to the most powerful AI tools may become a privilege of large corporations and government entities. This will limit the ability of independent developers and startups to test and implement cutting-edge technologies, forcing them to switch to less powerful or open models.
What Is Not Yet Known / Limitations
Based on current data, no explicit technical disagreements have been identified; the main discussions are focused on the socio-economic and infrastructural consequences of control.
Sources
Author
Look at AI, Editorial Staff