Anthropic has restored customer access to its powerful new AI model, Fable, after a more than two-week blackout imposed by US government safety concerns. The San Francisco-based company said export controls had been lifted following collaboration with the US Commerce Department.
The US government had ordered Anthropic to restrict access to Fable 5 and Mythos 5 on 12 June, banning use by foreign nationals over fears the models could enable devastating cyber-attacks. The ban was partially reversed on Friday, allowing Mythos 5 to be released to a small number of trusted US organisations for defensive cybersecurity.
US Commerce Secretary Howard Lutnick said in a letter to Anthropic that the company had agreed to proactively detect security risks, work with the government on protocols, and inform authorities of any malicious activity. Anthropic said it has increased safeguards and will scale up government collaboration.
The export controls followed a report by Amazon researchers who found a method to bypass Fable 5's safeguards to identify software vulnerabilities, potentially exploitable by malicious actors. Anthropic responded by temporarily suspending access entirely.
The Five Eyes intelligence agencies—Australia, the US, the UK, New Zealand, and Canada—warned last week that frontier AI models are anticipated to transform both offensive and defensive cyber capabilities within months, not years.
Criticism of the government's vetting process has emerged, with OpenAI CEO Sam Altman stating he does not like the idea of the government picking customers. OpenAI itself delayed the full public launch of GPT-5.6 at US government request.
