Anthropic's Powerful AI Models Reinstated After US Government Lifts Restrictions

The Trump administration has lifted export restrictions on Anthropic's advanced AI models, allowing the company to resume public access starting today. Howard Lotenick, the US Secretary of Commerce, stated that Anthropic collaborated closely with the government to address risks related to their Myth 5 and Fable 5 models. Over the past two weeks, they worked together to analyze and approve Fable 5, ensuring compliance with US regulations and reinforcing American leadership in AI.

Anthropic had suspended access to Myth 5 and Fable 5 on June 12, shortly after their release, due to a US export control order that barred foreign nationals, including those inside the US, from using these models. Since Anthropic could not distinguish between American and foreign users, it had to block access for all non-US citizens, including some employees involved in model development.

Myth was initially unveiled in April with advanced cybersecurity capabilities for scanning software code and identifying vulnerabilities. Due to concerns about misuse by malicious actors, access was limited to about 50 trusted organizations. Anthropic then developed Fable, a version with reduced cybersecurity functions and safeguards restricting use in high-risk areas such as biology and chemistry, as well as limitations on developing other large language models (LLMs). Fable launched just days before the Trump administration's export order.

The government’s block followed a report from Amazon researchers who demonstrated ways to bypass Fable’s protections to identify minor security flaws. Although Anthropic argued that weaker models could detect these flaws, the company enhanced safety mechanisms in close cooperation with the government. They claim the new safeguards block 99% of the risky behaviors described in the report, with only minimal, non-detailed information occasionally accessible.

According to the Wall Street Journal, the restrictions were lifted after Anthropic’s Chief Technology Officer, Tom Brown, engaged in discussions with the Commerce Department and other agencies to prove the model’s safety for public use. Anthropic expressed gratitude to users for their patience and to all partners involved in redeploying the models.