Anthropic reinstated customer access to Claude Fable 5 on July 1 across its controlled platforms, including the Claude API and Claude Code. The company added extra safety measures as part of an agreement with the U.S. Department of Commerce; some cybersecurity-related queries are now blocked or routed to the less capable Claude Opus 4.8.
What led to the suspension
Anthropic previewed Claude Mythos for select government and tech partners in April, then released Claude Fable 5 to customers worldwide on June 9. From the start, the models included guardrails: Fable 5 refused certain cybersecurity and biological-research queries and altered responses regarding how to build powerful AI models, which drew controversy.
The situation escalated after Amazon researchers used Claude Fable 5 to obtain information about how to carry out a cyberattack, prompting the U.S. Department of Commerce to issue an export-control directive. Anthropic has argued that any sufficiently capable model — including its own Claude Opus 4.8 and other providers’ models — could identify the same vulnerability and produce an exploit.
Export controls and partial redeployment
On June 12, the U.S. government issued an export-control directive suspending access to Claude Fable 5 and Mythos 5 for any foreign national, citing national-security concerns; Anthropic disabled access to the models globally the same day.
After negotiations, the government told Anthropic on June 26 that Claude Mythos 5 could be redeployed for select government organizations. In a letter, Commerce Secretary Howard Lutnick said the talks had produced "significant progress" and that Anthropic was committed to working with the government on protocols and standards for security assessments of AI models.
The export controls were lifted on June 30 after Anthropic implemented new safety guardrails designed to address the cybersecurity risk identified by the Amazon researchers. Anthropic announced Claude Fable 5 would be available worldwide beginning July 1; access for AWS, Google Cloud, and Microsoft Foundry customers followed shortly after.
User reports and access limits
In the days after Fable 5’s return, some users reported that the model’s performance had degraded compared with its earlier release: basic biological-science questions were being censored and coding tasks faced tighter restrictions. Anthropic said in an X post that some routine coding queries would fall back to Opus 4.8, and that it would try over the coming weeks to better distinguish genuine misuse from legitimate requests.
Other users complained that the model would soon be available for only 50 percent of a subscriber’s usage credits, with additional access purchasable per credit. Anthropic ultimately extended paid subscribers’ full Claude Fable 5 access through July 12.
Background and why this matters
Earlier in February, the Pentagon labeled Anthropic a "supply-chain risk" after the company declined to provide a version of its model without guardrails preventing use for mass surveillance or autonomous weapons; that designation barred the Defense Department and its partners from using Anthropic products. The June export-control order was the first instance where a government intervention led to suspension of general access to a model.
OpenAI’s recent rollout of new, powerful models in the GPT-5.6 family was also preceded by a government-mandated preview of the technology’s capabilities.
The episode marks a pivotal moment: governments are increasingly willing to scrutinize and, where they see fit, limit access to the most advanced AI systems before broad deployment. That approach affects how companies plan model releases and may prompt other countries to develop their own frontier models or seek less-restrictive partners to preserve technological sovereignty.
What to watch next
If governments intervene in AI model rollouts, there is a clear need for predictable, transparent, and fair processes. Companies require clarity on the standards they must meet to bring products to market; ad hoc decisions can delay innovation and erode public trust.
(This article summarizes the main milestones in the negotiations between Anthropic and the U.S. government, the suspension and reinstatement timeline, user reactions, and the broader industry implications.)
