.png)
8 hours ago
3
Anthropic has restored customer access to its powerful newest AI model, Fable, after a more than two-week blackout resulting from US government safety concerns that it could be abused to enable serious cyber-attacks.
The San Francisco company said export controls had been lifted, citing a social media post from the US commerce secretary, Howard Lutnick, in which he said: “Over the past two weeks, we have worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the US Government and strengthen America’s leadership in AI.”
Three days after Claude Fable 5 was launched on 9 June, the US government ordered Anthropic to immediately restrict access to it and a related model, Claude Mythos 5, banning their use by foreign nationals, whether inside or outside the US. Washington feared the powerful AIs could be used to mount devastating cyber-attacks by foreign adversaries.
On Friday the US government allowed Anthropic to release Mythos 5, to some “trusted” US organisations, partly reversing the order. Mythos was only released to a small number of trusted partners for use in defensive cybersecurity.
Lutnick said in a letter to Anthropic, seen by Reuters: “Anthropic has agreed to proactively detect and address security risks associated with the models; to work diligently with the US government on protocols and standards and releases for Mythos, Fable, and future models; and to inform the US government of any malicious activity.”
The ban was seen as a toughening in the White House’s stance towards the regulation of US AI companies, which had previously emphasised a light touch underpinned by a desire to beat China in the supposed race to AI dominance.
The US government acted after it became aware of a report in which Amazon researchers found a method of bypassing Fable 5’s safeguards to make it find software vulnerabilities, showing how such a weakness could be exploited – potentially by a malign cyber-attacker.
Anthropic responded to the export control by pulling the plug entirely, but it said it has now increased safeguards. It said in a blogpost: “Researchers from the US Department of Commerce’s Center for AI Standards and Innovation (CAISI) have tested both our prior and new safeguards and agree that they are extraordinarily strong.”
Anthropic said late on Tuesday it would “scale up our government collaboration” and “stand up dedicated Anthropic teams to work on shared government priorities”.
Last week intelligence agencies in Australia, the US, the UK, New Zealand and Canada warned that, while AI “would help us improve cyber defence over time, it also accelerates the speed, scale, and sophistication of cyber threats”.
The warning by Five Eyes agencies said: “Frontier AI models are anticipated to exceed current industry expectations, fundamentally transforming both offensive and defensive cyber capabilities. The timeline is not years, it is months.”
after newsletter promotion
Susie Wiles, Donald Trump’s chief of staff, posted: “My gratitude to companies across industries who continue to work closely with the White House to implement the President’s EO: ‘Promoting Advanced AI Innovation and Security.’ This includes excellent work around advanced model access and guardrail testing and security.
“The government and private sector have worked together in a way we have never seen before and this foundation of America First is unprecedented. Our shared priority remains: get the best tech deployed as quickly and safely as possible.”
But the US government’s vetting of which companies can gain access to the models has drawn criticism.
The OpenAI chief executive, Sam Altman, said last week that extensive safety testing “is not a bad idea. I just don’t like the idea of the government picking the customers.” The ChatGPT maker delayed a full public launch of GPT-5.6 at the US government’s request, limiting its access to a small group of vetted partners.
