Frontier AI access becomes strategic privilege under US government oversight (legrandcontinent.eu)

• OpenAI rolls out GPT-5.6 slowly, initially only to approved US companies.
• Anthropic restores Claude Mythos 5 for US critical infrastructure operators.
• US considers banning Chinese open-source AI models and tightening export controls.
• China may either adopt licensing or continue open source; latter threatens US labs.
• Europe must choose between US alignment or €500 billion autonomous investment.

"Access to frontier AI models is increasingly controlled by the US government. OpenAI and Anthropic are distributing their latest models only to approved US entities. China faces a choice between adopting a similar licensing regime or continuing open-source releases, which could disrupt US business models. Europe must decide between integration with the US or a costly autonomous path."