U.S. reportedly asks OpenAI and Anthropic to withhold new models from UK testers
A post sharing a Politico report says Anthropic complied, keeping Mythos 5.1 U.S.-only, while the UK still tested OpenAI’s GPT-6 Astra.
TLDR
A post sharing a Politico report says the U.S. government asked OpenAI and Anthropic to withhold new frontier models from the UK AI Security Institute until a U.S. review was complete. It says Anthropic complied, keeping Mythos 5.1 U.S.-only, while OpenAI’s GPT-6 Astra was still tested by the UK. The post links the request to cybersecurity fears and describes it as administrative pressure, not a formal export ban.
U.S. reportedly asks OpenAI and Anthropic to withhold new models from UK testers
A post sharing a Politico report says Anthropic complied, keeping Mythos 5.1 U.S.-only, while the UK still tested OpenAI’s GPT-6 Astra.
TLDR
A post sharing a Politico report says the U.S. government asked OpenAI and Anthropic to withhold new frontier models from the UK AI Security Institute until a U.S. review was complete. It says Anthropic complied, keeping Mythos 5.1 U.S.-only, while OpenAI’s GPT-6 Astra was still tested by the UK. The post links the request to cybersecurity fears and describes it as administrative pressure, not a formal export ban.
