Washington Removes Export Restrictions on Anthropic's Fable 5 and Mythos 5 AI Models

The U.S. government has officially removed export controls on two of Anthropic's most advanced AI models — Claude Fable 5 and Mythos 5. The Commerce Department formally cleared both systems on June 30, allowing the company to begin restoring full global access starting July 1. The move brings an end to nearly three weeks of intense negotiations between Anthropic and federal officials.

Anthropic confirmed the news via a post on X: "We've received notice that the Department of Commerce has lifted export controls on Claude Fable 5 and Mythos 5. We'll begin restoring access tomorrow, and will share an update soon. We're grateful to our users for their patience, and to everyone who worked with us on this."

**How the Export Controls Came Into Effect**

Export controls are federal regulations that limit access to sensitive technologies — including advanced AI systems — on national security grounds. In this case, the Commerce Department issued its directive on June 12, just three days after both models launched on June 9. Officials cited national security concerns reportedly tied to potential vulnerabilities related to jailbreaking the models.

Following the directive, Anthropic was required to suspend access for foreign nationals across all platforms. However, segmenting users in real time by nationality turned out to be practically unworkable. As a result, the company had no choice but to take both models entirely offline — affecting users on Claude.ai, the API, AWS Bedrock, and all partner integrations.

The situation sparked immediate disruption for businesses, developers, and enterprise customers relying on the models. With both systems down, entire workflows were halted until a resolution could be reached.

**What Changed to Lift the Ban**

Over the roughly three-week standoff, Anthropic engaged in direct, intensive talks with Commerce Department and White House officials. Commerce Secretary Howard Lutnick confirmed the outcome on June 30, stating: "Over the past two weeks, we have worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the U.S. Government and strengthen America's leadership in AI."

The swift resolution is being seen as both a win for Anthropic's advocacy efforts and a sign that regulatory frameworks around advanced AI are becoming more responsive and sophisticated.

**Why These Models Are Significant**

Claude Fable 5 is Anthropic's most capable and widely deployed model available to the general public. Built on the company's powerful Mythos-class architecture, it comes equipped with enhanced safeguards tailored for broad use. The model performs exceptionally well on complex reasoning tasks, long-horizon agentic workflows, software engineering challenges, and advanced visual processing.

Mythos 5, by contrast, is designed for more sensitive, high-stakes environments. While it shares the same underlying architecture as Fable 5, it incorporates additional safeguards specifically for cybersecurity applications. Prior to the export control dispute, access to Mythos 5 was limited to trusted partners through a dedicated initiative known as Project Glasswing, primarily serving government and enterprise deployments.

**Pricing and Safety Features**

Anthropic has positioned both models competitively within the AI industry. The company charges $10 per million input tokens and $50 per million output tokens. Both models also feature built-in classifiers that automatically redirect high-risk queries to safer fallback responses, particularly for tasks touching on cybersecurity and biology.

The resolution of this dispute underscores a broader trend: as AI capabilities advance rapidly, government regulators and AI developers are being forced to find workable frameworks that balance national security concerns with the practical realities of global technology deployment. For Anthropic, the episode also demonstrates that proactive engagement with federal agencies can produce swift results — even in politically sensitive situations.