Anthropic Suspends Claude Mythos and Fable at U.S. Request

2026-06-13

Anthropic Halts Claude Mythos and Fable at U.S. Request

Immediate suspension of services

U.S. AI startup Anthropic said on the 12th that it would immediately stop offering its advanced AI models 'Claude Mythos 5' and 'Claude Fable 5'. The company said the U.S. government had designated them as subject to export controls and asked that foreign nationals be prevented from using them.

Consumer apps and enterprise technology services were suspended as of the 12th. The general rollout began on the 9th, meaning the models had been public for only a short time.

Unusual government designation

It is unusual for AI models to be included in export controls as software technology. So far, the U.S. government has focused on tightening restrictions on hardware used in AI development, such as semiconductors. The move marks a further step in treating AI as a strategic resource tied to national security.

On June 2, Anthropic expanded access to 'Mythos', which had been centered on the U.S., to more than 15 countries including Japan. In Japan, major banks and others had obtained access rights to 'Claude Mythos Preview'. It is not clear whether Japanese companies or the government have suspended use.

Request to halt access for all foreign nationals

According to the company, it received a letter from the U.S. government at just after 5 p.m. U.S. Eastern time on the 12th, or just after 6 a.m. Japan time on the 13th. The letter asked the company, for national security reasons, to block foreign nationals from accessing Mythos 5 and Fable 5.

The letter sought to stop provision to all foreign nationals, including Anthropic employees with foreign citizenship and residents in the U.S. In response, the company halted access for all customers, including U.S. citizens. According to multiple U.S. media reports, employees with foreign citizenship also cannot use the models.

Anthropic said the U.S. government may have designated Fable for export controls over concerns that it could have a safety loophole. The company pushed back, saying there were no serious safety issues, and said: 'We think this is a misunderstanding and we are working to restore access as soon as possible.'

Concern over business impact

Anthropic said the U.S. government had not provided a specific explanation of any national security risk. 'If similar standards were applied across the industry, the release of new models by advanced model providers would effectively stop,' it said, criticizing the government response as excessive.

In April, the company unveiled 'Mythos', an AI said to excel at finding system vulnerabilities. It had withheld a public release over concerns it could be misused for cyberattacks. In the early stage, it was made available only to 50 companies and organizations, mainly U.S. firms, and later access requests came in from Japan and other countries.

Meanwhile, Fable, which has comparable performance, was released after additional safety measures were strengthened.

Enjoyed this article? Share it with your network!