Trump Administration Reopens Access to Anthropic's Mythos 5 Model

A Temporary Ban Leads to Compromise

In mid-June 2025, Anthropic withdrew two of its most powerful cybersecurity models from the market — Mythos 5 and Fable 5 — following a government ban imposed by the Trump administration. Within just two weeks, however, signs of a reversal began to emerge as the administration opened the door to negotiations, resulting in a partial restoration of access to these models.

What Changed?

U.S. Secretary of Commerce Howard Lutnick sent a formal letter to Anthropic's Chief Computing Officer, Tom Brown, announcing that "appropriate safeguards are now in place to allow trusted partners to access the Claude Mythos 5 model." Under this decision, the model is made available to more than 100 designated U.S. government agencies and American companies, including non-U.S. employees working within those organizations, as well as non-American Anthropic employees who had been caught under the original ban.

Why Was the Ban Issued in the First Place?

The roots of this issue lie in a critical security problem. Fable 5 had been released just days before the ban as a supposedly safer version of Mythos 5, yet cybersecurity researchers were able to bypass the built-in safeguards of both models with remarkable ease. This raised serious concerns on the government's part and prompted an immediate blanket ban.

What Remains Banned?

Notably, the new decision was limited to reinstating Mythos 5 and made no mention of Fable 5. Despite Fable 5 having originally been designed for general use, it remains blocked pending a separate ruling on its status.

Anthropic's Official Position

The company publicly acknowledged the recent developments via X, clarifying that it had been working with the government since June 12 to restore access to both models. It noted that Mythos 5 would be redeployed to serve American organizations that operate and defend critical infrastructure, and confirmed that it is seeking to gradually expand access to include Fable 5 as well.

Broader Implications

This incident reflects a growing tension between the imperatives of technological innovation and the demands of national security in the age of artificial intelligence. It highlights a number of fundamental challenges:

• The difficulty of adequately controlling and securing AI models specialized in cybersecurity before their release.
• The expanding role of the U.S. government in regulating who has access to advanced models.
• The need for more rigorous verification mechanisms before releasing models with potentially offensive capabilities.
• Questions about whether companies can fulfill their security commitments under the pressures of competition and rapid deployment.

The Fable 5 file remains open, and how it is ultimately handled will set an important precedent for how the U.S. administration approaches the future regulation of highly sensitive AI models.

✦ بقلم فريق دروب أيديا