US set to restore Anthropic's Fable 5 AI model after security review: Report

The Donald Trump administration is nearing a decision that could allow Anthropic to restore public access to its flagship Fable 5 artificial intelligence model, according to an Axios report.

The Fable 5 model by Anthropic has remained unavailable for the past 15 days after the US government raised national security concerns, disrupting developers and businesses that had begun using it shortly after launch.

Axios, citing a officials familiar with the discussions, reported that restrictions on Fable 5 could be lifted as early as next week. Another official told the publication that negotiations are expected to continue over the weekend and that Anthropic is hopeful of restoring access soon.

Government review appears to be entering final stages

While optimism has grown around Fable 5's return, the model has not yet received clearance from every agency involved in the review. According to Axios, the Pentagon and the National Security Agency still need to approve the release, meaning the outcome remains uncertain despite positive developments elsewhere within the administration.

Other US government departments have concluded the model, Fable 5,can safely return to public use, signalling a significant shift after months of tense discussions between Anthropic and federal officials.

Commerce Secretary Howard Lutnick and Treasury Secretary Scott Bessent have reportedly played key roles in easing the dispute between the administration and the AI company.

Commerce Department restores limited access to Mythos 5

In a related development, the Commerce Department on Friday allowed Anthropic to restore access to Mythos 5, its most advanced cybersecurity-focused AI model, for a select group of trusted users.

Unlike Fable 5, Mythos 5 has never been broadly available and includes safeguards intended to reduce the risk of misuse in cyberattacks or biological threats.

In a letter to Anthropic, Lutnick wrote that the company "has worked with the US government to address risks associated with" Mythos 5 and Fable 5.

"These efforts," he continued, "have yielded significant progress. In addition, Anthropic has committed to work with the US government on protocols and standards and releases."

Fable 5 became a favourite among developers before suspension

Before access was suspended on June 12, Fable 5 had attracted strong interest from software developers and early adopters because of its advanced reasoning and coding capabilities.

Many developers regarded it as a significant step forward for AI-assisted programming. The report noted that newer models, including open-source alternatives, have increasingly been judged against Fable 5's performance.

Anthropic had briefly made the model available at no additional charge for users on several paid Claude subscription plans until June 22, allowing subscribers a limited opportunity to test its capabilities before access was withdrawn.

The publication also highlighted early testing by payments company Stripe, which reportedly used Fable 5 to overhaul a 50 million-line codebase in a single day, a task that would otherwise have taken engineers more than two months to complete manually.

Following the suspension, developers reportedly saw automated coding tasks abruptly halted, while some businesses shifted workloads to competing AI systems, including lower-cost Chinese models.

Tensions between Anthropic and Trump admin

The expected return of Fable 5 marks a notable easing of tensions after a four-month dispute between Anthropic and the Trump administration over national security concerns.

One administration official told Axios that Anthropic “has worked positively with the government.”

The latest remarks contrast with the administration's earlier criticism. Following disagreements over how the Pentagon could use Claude, Defence Secretary Pete Hegseth had previously labelled Anthropic a "Supply-Chain Risk to National Security," .

AI companies seek a clearer approval framework

Axios reported that both Anthropic and OpenAI are pressing the Trump administration to establish a formal process for reviewing advanced AI models before release.

The discussions follow President Donald Trump's June 2 executive order, which introduced a framework for voluntary government vetting of the most powerful AI systems.

When Anthropic suspended Fable 5 and Mythos 5 on June 12, the company said it wanted "a statutory process that is transparent, fair, clear, and grounded in technical facts," adding that the administration's decision to restrict the two models "does not adhere to those principles."

OpenAI echoed similar concerns after receiving permission on Friday to begin a limited preview of GPT-5.6.

In a blog post OpenAI said: “We don't believe this kind of government access process should become the long-term default. It keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them.”