Anthropic sends staff to Washington to lift its model export ban
Anthropic has sent senior staff to Washington to negotiate the removal of an export-control directive that forced it to disable its Fable 5 and Mythos 5 models worldwide, The Hill reports . The administration invoked national-security authority on June 12 to bar foreign governments, companies and individuals from using the two models — a restriction broad enough to cover Anthropic’s own non-U.S. employees. The company’s delegation reportedly included safety and security leads alongside co-founder Tom Brown and policy chief Sarah Heck, meeting officials including the commerce secretary and the national cyber director.
The dispute turns on how serious the underlying finding is. The restriction stems from research by Amazon scientists who reported that Fable 5’s safeguards could be partially bypassed, which alarmed White House officials. Anthropic says it reviewed what it believes is that report and concluded the demonstrated capability is no greater than what is already available from other publicly deployed systems, naming OpenAI’s GPT-5.5. That gap matters: if regulators treat one company’s model as a controlled item while comparable models stay freely available, the control does little for security while cutting the affected firm off from overseas customers and staff — the commercial stakes driving Anthropic’s push for a quick deal.