US government order forces Anthropic to halt advanced AI model distribution
The US artificial intelligence startup stops distribution of models noted for high vulnerability identification capabilities in compliance with federal mandate.
Anthropic, a US-based artificial intelligence startup, has announced the immediate cessation of distribution for its advanced AI models, Claude Mutus 5 and Fabulous 5. The decision, made public on 12 June 2026, was executed in direct compliance with an order issued by the US government.
The affected models are characterised by their exceptional capability in identifying software vulnerabilities. This technical proficiency has drawn significant attention within the cybersecurity sector, with previous reports indicating that major Japanese security firms had already adopted the Claude Mutus-class AI for infrastructure defence purposes.
While the specific legal or regulatory basis for the US government’s directive remains unspecified in available reports, the company confirmed it is acting under a mandatory order. The precise nature of the directive, whether it constitutes an injunction, executive order, or regulatory measure, has not been detailed by official sources.
The scope of the distribution halt is also not explicitly confirmed in the source material. It remains unclear whether the restriction applies globally or is limited to US jurisdiction, although the announcement implies a broad cessation of availability for these specific models.
This development follows the general availability of the Claude Mutus-class AI, which was reported to have been released to the public earlier in June. The sudden halt raises questions regarding the long-term implications for Anthropic’s business operations and the future release of these high-capability models, though no further details on these aspects were provided.
NHK News Japan reported the announcement, noting the models' extremely high capability in vulnerability detection. The source treats this description as a claim regarding the models' performance rather than an independently verified technical metric.
The event underscores the increasing intersection between federal governance and artificial intelligence development. As governments seek to regulate powerful computational tools, the compliance requirements imposed on private sector entities like Anthropic are likely to shape the deployment of advanced AI systems in critical infrastructure sectors.
