Cyber leaders urge US to lift curbs on Anthropic's security models

The US is under pressure to lift emergency export controls on Anthropic's Fable 5 and Mythos 5, with cyber leaders arguing the limits hinder defenders more than attackers. An open letter dated June 15, 2026 and signed by more than 50 executives from Nvidia, Adobe and other tech firms arrived two days after the Commerce Department imposed the restrictions following a reported jailbreak that bypassed safety guardrails. Anthropic responded by globally suspending access to the two models, saying nationality-based enforcement was impractical. Signatories contend the restrictions offer little extra safety beyond existing platforms and could slow industry progress. The controversy sits at the intersection of national security, model safety standards, and industry vitality, amid broader friction with Washington over Anthropic’s safety posture and a Department of Defense designation as a supply-chain risk. With Anthropic valued around $965 billion and hints of an IPO, the timing matters as policy and frontier AI development collide, including initiatives like Project Glasswing.