Claude Fable 5 access is being restored after the U.S. Department of Commerce lifted export controls that had forced Anthropic to suspend two of its newest AI models.
Anthropic said the restrictions on Claude Fable 5 and Claude Mythos 5 were lifted on June 30, clearing the way for the company to bring Fable 5 back to users globally starting July 1. The company said access to Mythos 5 has also been restored for a set of approved U.S. organizations, with broader expansion planned through its Glasswing program.
The reversal ends a disruption that began June 12, when U.S. officials ordered Anthropic to restrict access to Fable 5 and Mythos 5 for foreign nationals, including non-U.S. Anthropic employees. Because the order took effect immediately and the company said it could not verify nationality in real time, Anthropic disabled both models for all users.
The episode is one of the clearest signs yet that frontier AI releases are moving into a more regulated era, especially when national-security officials believe advanced models could be misused for cyber operations.
Claude Fable 5 Returns After Export Controls Are Lifted
Anthropic said Fable 5 will return to the Claude Platform, Claude.ai, Claude Code and Claude Cowork. The company also said it would re-enable access through AWS, Google Cloud and Microsoft Foundry as quickly as possible.
For some paid users, Fable 5 will be included within weekly usage limits through July 7, after which access will shift to usage credits, according to Anthropic. That gives customers a short transition period as the company restarts access after nearly three weeks of disruption.
The company has not framed the restoration as a simple return to normal. Instead, Anthropic said it added safeguards, expanded government coordination and began work with major technology partners on a common framework for evaluating AI jailbreak risks.
Reuters reported that the Commerce Department lifted the restrictions less than three weeks after the original order, following Anthropic’s implementation of new safeguards.
Why the Models Were Suspended
The June 12 directive centered on national-security concerns.
Anthropic said at the time that the government ordered it to suspend access to Fable 5 and Mythos 5 for any foreign national, whether inside or outside the United States. The company said the directive did not provide detailed written specifics on the security concern, but Anthropic understood it related to a reported method of bypassing Fable 5’s safeguards.
The company said it reviewed a demonstration tied to a small number of previously known and relatively simple software vulnerabilities. Anthropic argued that comparable information could be produced by other publicly available models and said the issue did not justify recalling a commercial model used at large scale.
After the shutdown, Anthropic said the government’s concern had been triggered by a report from Amazon researchers. The company said it then worked with government officials and partners, including Amazon, to review the evidence and adjust its safeguards.
New Safeguard Blocks Reported Bypass
Anthropic said it trained an improved safety classifier to block the behavior described in the report that led to the order.
The company said the new safeguard blocks the specific technique in more than 99% of cases. Requests blocked by the safeguard will be routed to Anthropic’s Opus 4.8 model instead, according to the company.
That change may come with trade-offs. Anthropic said the new classifier could flag more benign coding and debugging requests, increasing false positives for some users. The company said it would continue refining the system to better distinguish legitimate use from misuse.
For enterprise customers and developers, that means Fable 5 may be more available than it was during the suspension, but not necessarily less restrictive in cybersecurity-related workflows.
Mythos 5 Remains More Limited
The return of Mythos 5 is narrower than the return of Fable 5.
Anthropic said Mythos 5 has been restored for a set of U.S. organizations after government approval on June 26. The company said it is still coordinating with U.S. officials to expand access to domestic and international partners in the Glasswing program.
That limited rollout reflects the different role of Mythos 5. Anthropic has described Mythos 5 as a model intended for defensive cybersecurity work with trusted partners, while Fable 5 is the more broadly available model with stronger safeguards.
The split underscores a likely direction for frontier AI products: broad public models may face one set of safeguards, while more capable security-focused models may remain restricted to vetted organizations.
U.S. Oversight of AI Models Is Tightening
The dispute also shows how quickly AI regulation can affect global access.
Reuters reported that Washington has increased oversight of new model releases amid concerns that advanced AI systems could be misused by foreign military or intelligence actors. The original order required Anthropic to restrict access by foreign nationals, prompting the company to shut down both models worldwide because it could not meet the requirement instantly.
Reuters also reported that Commerce Secretary Howard Lutnick told Anthropic the department could reevaluate the decision and reimpose a license requirement if circumstances change or if Anthropic fails to meet its commitments.
That warning keeps regulatory pressure on the company even after the immediate restrictions were lifted.
Industry Standards May Become the Next Battleground
Anthropic said it is working with Amazon, Microsoft, Google and other Glasswing partners to develop shared standards for assessing and fixing AI jailbreaks. The company said a consistent framework would help developers rank the severity of reported bypasses and respond more predictably.
That could become important for the entire AI industry. If governments judge model safety case by case without a shared technical standard, companies may face sudden access restrictions, uneven enforcement and uncertainty around product launches.
For Anthropic, the Fable 5 case has already become more than a product-access issue. It is now a test of how AI companies, cloud partners and regulators coordinate around frontier model safety.
The next thing to watch is whether Fable 5’s restored rollout remains stable, whether Mythos 5 access expands beyond approved U.S. organizations, and whether the U.S. government turns this ad hoc intervention into a more predictable approval system for advanced AI models.
Read Also: NTSA Instant Fines: How Kenyan Drivers Pay Tickets
