Two weeks ago, Anthropic pulled the plug. The company disabled global access to its most powerful cybersecurity model, Mythos 5, and a second model called Fable 5. The order came from Washington: no foreign nationals could use them.
The government feared security guardrails could be bypassed. Now, some of that access is back.
The Commerce Department has cleared Anthropic to restore Mythos 5 to a small, select group of cyber defenders and infrastructure providers. Commerce Secretary Howard Lutnick wrote to Anthropic’s chief compute officer that the company’s work with the government “have yielded significant progress.” The door cracked open. But not all the way.
Fable 5 remains locked down. That model still cannot answer certain queries — specifically questions tied to cybersecurity and biology.
When users ask, Anthropic’s Claude chatbot routes those responses through other channels. The recent letter from Lutnick made no mention of easing restrictions on Fable 5. It stays under the earlier order.
Anthropic is now working to provision the approved providers and restore their access quickly. The company said it is optimistic and will continue working with the government to expand access to Mythos 5. It also wants to make Fable 5 available for general use again.
No timeline was given. The speed of the reversal is notable.
Two weeks. That is how long it took for Anthropic to address national security concerns to the government’s satisfaction. “In just two weeks, we have worked diligently to ensure America remains the global leader in AI while safeguarding our security,” said Benno Kass, a Commerce Department spokesman.
This is not a full restoration. It is a limited one.
A small group of trusted partners gets Mythos 5 back. The rest of the world waits. The government still holds the keys.
And the underlying tension — between AI capability and national security — remains unresolved. Mythos 5 is Anthropic’s strongest cybersecurity model. It was designed to defend networks, find vulnerabilities, and respond to threats.
But the same power that makes it effective also makes it risky. The government worried that adversaries could exploit it.
That is why the order came down so abruptly two weeks ago. That is why Anthropic disabled access globally, not just for foreign nationals. The company entered intensive talks with the government.
The result is the current compromise: a small group of vetted partners gets access. The broader question — who else can use it, and under what conditions — remains open.
Benno Kass framed the outcome as a win for American leadership. Anthropic framed it as progress. Both sides avoided the word “ban.” Neither side claimed victory.
The model is back in limited use. The restrictions on Fable 5 stand. The talks continue.
For now, the cybersecurity community has partial access to a powerful tool. For now, the government keeps its hand on the lever.
For now, that is the deal.
{“@context”: “https://schema.org”, “@type”: “NewsArticle”, “headline”: “Commerce Department approves limited restoration of Anthropic’s Mythos 5 AI”, “description”: “Two weeks ago, Anthropic pulled the plug. The company disabled global access to its most powerful cybersecurity model, Mythos 5, and a second model called…”, “datePublished”: “2026-06-27T07:07:31+00:00T09:30:00+00:00”, “author”: {“@type”: “Person”, “name”: “Isabella Moore”}, “publisher”: {“@type”: “Organization”, “name”: “InfoPulse Today”, “url”: “https://newsfeed.infopulsetoday.com/”}}
