US Government Lifts AI Export Controls on Anthropic While Restricting OpenAI, Setting New National Security Precedent

The US Commerce Department lifts export controls on Anthropic's AI models after the company agrees to government safety protocols, while simultaneously restricting OpenAI's GPT-5.6 launch over national security concerns, marking a landmark shift in how Washington regulates powerful AI technology.

• The US Commerce Department lifts export controls on Anthropic's Mythos and Fable AI models after Anthropic agrees to proactively detect security risks, collaborate with the government on safety protocols, and report any malicious activity.
• The US government is also restricting OpenAI's GPT-5.6 to a limited preview launch, citing similar national security concerns about powerful AI being exploited for cyberattacks or military use by adversarial nations.
• Washington's interventions are setting a new precedent for AI oversight, sparking calls for formal pre-release review frameworks and safety standards comparable to those applied to nuclear technology.