US Approves Claude Export
Anthropic's Claude 5 models are officially cleared for global release after a national security freeze! 🔓 Here is what changed.
What the video says
Just 3 weeks after a national security scare froze their deployment, Anthropic's flagship Claude Fable-5 and Mythos-5 models have had their export curbs lifted by the US government for global release. This decisive pivot comes after the release of Mythos spooked Trump into requesting voluntary safety testing in May, triggering a temporary shutdown of advanced model access.
To resolve the tension, US Commerce Secretary Howard Lutnick confirmed Anthropic no longer needs a license for exports, marking a landmark a key change in how Washington regulates Frontier AI. In exchange, Anthropic expanded its government partnership and established a dedicated team monitoring jailbreak threats 24/7 to ensure 99% of known exploits are blocked.
However, this unprecedented coordination introduces a significant trade-off, as Anthropic warns that Fable-5's tightened safety safeguards may now block benign prompts during routine coding tasks. While White House Chief of Staff Suzy Wiles hailed the private-public cooperation as unprecedented, Anthropic CEO Dario Amodei warns that the government is still moving too slowly to establish formal safety standards.
Ultimately, the Frontier Security Institute suggests the biggest question mark is how the U.S. administration will handle equivalently dangerous capabilities coming from China, with fewer guardrails.
