The U.S. Department of Commerce has lifted export controls on Anthropic's Claude Fable 5 and Mythos 5 AI models, allowing the company to restore access starting Wednesday. The reversal comes less than three weeks after the government ordered the suspension of these models over national security concerns, citing potential risks of exploitation by hackers.
Part 1: Immediate Action & Core Facts
The Commerce Department notified Anthropic on Tuesday that export controls had been lifted, enabling the company to resume access to its advanced AI tools. The suspension, imposed on June 12, barred foreign nationals—including Anthropic employees—from accessing the models. The reversal follows a partial lifting of restrictions last week, which allowed Mythos 5 to be shared with select U.S. organizations.
Part 2: Deeper Dive & Context
Background on the Suspension
Anthropic disabled access to Fable 5 and Mythos 5 after the Commerce Department cited national security risks, particularly concerns that the models could be exploited to bypass software safety restrictions—a process known as "jailbreaking." The company disputed the government's assessment, arguing that the suspension was based on a narrow potential vulnerability rather than a systemic flaw.
Government Rationale
Commerce Secretary Howard Lutnick stated that the government had worked with Anthropic to ensure the models aligned with U.S. AI leadership goals. The initial suspension coincided with rising competition from Chinese AI models, which some tech executives warned could gain an advantage during the U.S. restrictions.
Anthropic's Response
Anthropic expressed gratitude to users and partners for their cooperation during the suspension. The company had previously sued the Trump administration over a separate designation as a supply chain risk, a legal battle that remains unresolved. The reversal comes as Anthropic prepares for a potential IPO later this year.
Model Capabilities
Fable 5 is designed for consumer use, capable of deep reasoning and complex tasks. Mythos 5, intended for businesses and cybersecurity experts, can identify and exploit vulnerabilities in computer code. The models are part of Anthropic's Claude platform, a competitor to OpenAI's ChatGPT and Google's Gemini.
Broader Implications
The suspension and subsequent reversal highlight ongoing tensions between AI companies and U.S. regulators over balancing innovation with national security. The episode also underscores the Trump administration's efforts to maintain U.S. leadership in AI amid global competition.