Anthropic Restores Fable Access After U.S. Controls Lifted

Anthropic has begun restoring access to its Fable 5 AI model after the U.S. government lifted export controls that had forced the company into a sudden global shutdown.

The San Francisco AI company said the restrictions on Claude Fable 5 and Claude Mythos 5 were lifted on June 30, less than three weeks after the U.S. government ordered access blocked for foreign nationals.

Anthropic said the order originally covered foreign nationals both inside and outside the United States, including foreign-national employees working at the company.

Anthropic Restores Fable Access

In a company update, Anthropic said Fable 5 would become available globally from Wednesday, July 1, across Claude Platform, Claude.ai, Claude Code and Claude Cowork.

The company said it would also work to restore access through AWS, Google Cloud and Microsoft Foundry as quickly as possible.

Mythos 5, the more restricted cybersecurity-focused model, has been restored for a set of U.S. organisations following government approval on June 26.

Anthropic said it is continuing to coordinate with the government to expand access to more domestic and international partners through its Glasswing programme.

The result is not a simple return to normal service.

It creates a new access model where frontier AI releases are tied more closely to government testing, safety commitments and approved partner channels.

📰 Read Also: KPMG AI Hallucinations Warning Raises Audit Risk

Export Controls Followed Cybersecurity Fears

The restrictions began on June 12, when the U.S. government issued an export-control directive covering Fable 5 and Mythos 5.

At the time, Anthropic said the directive cited national-security authorities and required the company to suspend access by foreign nationals.

Because the order took effect immediately, Anthropic said it had no reliable way to verify nationality in real time and therefore disabled the models for all customers.

The company later said the concern followed a report in which Amazon researchers found a way to bypass Fable 5 safeguards.

That bypass involved getting the model to identify software vulnerabilities, including one case where it produced code showing how a vulnerability could be exploited.

Anthropic said it worked with the government, Amazon and other partners over the past two weeks to review the evidence and strengthen safeguards.

📰 Read Also: Dutch AI Image Ruling Sets Copyright Warning

Safeguards Become the Main Issue

Anthropic said researchers from the U.S. Department of Commerce’s Center for AI Standards and Innovation tested both the earlier and updated safeguards.

The company said the new classifier is designed to block the behaviour described in the Amazon report.

Blocked requests to Fable 5 will be redirected to Claude Opus 4.8, and users will be notified when that happens.

That detail matters for customers because the model may return to service with more visible safety routing than before.

For developers, the practical question is how often ordinary cybersecurity, coding or debugging work is blocked by safety filters.

For regulators, the deeper question is whether companies can safely release models with strong cyber capabilities before misuse risks are fully understood.

📰 Read Also: Amazon Zoox Recall Raises Robotaxi Safety Questions

AI Policy Enters a New Phase

The Fable and Mythos case shows how quickly frontier AI access can become a national-security matter.

The U.S. government did not permanently block the models, but it forced a pause that affected customers, cloud partners and Anthropic’s own deployment plans.

Anthropic now says it will deepen collaboration with Washington on pre-release testing, information sharing and research priorities.

It also says it is working with Amazon, Microsoft, Google and other Glasswing partners on a shared framework for judging the severity of AI jailbreaks.

That framework could become important far beyond this single dispute.

If governments begin treating advanced AI models like sensitive dual-use technology, model launches may increasingly depend on security reviews before full access is restored.

For customers, the lesson is direct: access to the most capable AI systems can change quickly when regulators decide a model creates cyber, national-security or export-control risk.

TL;DR

• Anthropic says U.S. export controls on Claude Fable 5 and Claude Mythos 5 have been lifted.

• The restrictions were imposed on June 12 and covered foreign nationals inside and outside the U.S.

• Anthropic disabled access for all users because it could not verify nationality in real time.

• The controls followed concerns about a reported bypass of Fable 5’s cybersecurity safeguards.

• Fable 5 is being restored globally, while Mythos 5 remains limited to approved organisations.

• The case may shape how future frontier AI models are tested, released and monitored.

Read More

• KPMG AI Hallucinations Warning Raises Audit Risk

• Dutch AI Image Ruling Sets Copyright Warning

• Amazon Zoox Recall Raises Robotaxi Safety Questions