Two weeks ago, the US government slammed the brakes on Anthropic’s most powerful cybersecurity AI. The order was blunt: bar foreign nationals from accessing Mythos 5 and its sibling model, Fable 5. The fear was straightforward — security guardrails could be bypassed.
Anthropic shut down global access the same day. Now, a partial reversal is underway.
Commerce Secretary Howard Lutnick sent a letter to Anthropic’s chief compute officer. The company’s work with the government, Lutnick wrote, “has yielded significant progress.” The result is narrow but concrete. Mythos 5 can return to a small group of cyber defenders and infrastructure providers.
Anthropic says it is working to provision those approved partners and restore their access as quickly as possible. What remains locked down carries its own weight.
Fable 5 stays under the original restrictions. That model cannot answer queries on cybersecurity or biology. When users ask, Anthropic’s Claude chatbot routes those responses through other channels.
Lutnick’s letter does not mention any change to Fable 5’s status. It remains frozen. The stakes here are not abstract.
Mythos 5 is Anthropic’s strongest cybersecurity model. In the hands of trusted defenders, it can scan networks, find vulnerabilities, and automate threat response.
That capability matters. Ransomware attacks hit hospitals, pipelines, and school systems. Critical infrastructure operators need every tool they can get.
The government’s concern was that the same power, if accessed by adversaries or careless foreign nationals, could be turned against US systems. The two-week suspension created a gap.
Defenders lost access to a tool they had come to rely on. Commerce Department spokesman Benno Kass put it plainly. “In just two weeks, we have worked diligently to ensure America remains the global leader in AI while safeguarding our security.” The timeline is tight.
Anthropic and the government negotiated a resolution in fourteen days. That is fast for any federal action. It suggests urgency on both sides.
Anthropic expressed optimism about the progress. The company said it will work with the government to expand access to Mythos 5 and to make Fable 5 available for general use again.
No timeline was given. The path forward depends on continued cooperation and monitoring. The government retains the power to reimpose restrictions if concerns resurface.
For now, the balance tilts back toward deployment. A small group of cyber defenders gets their tool back.
Infrastructure providers can harden their systems with AI assistance. But the episode reveals a deeper tension. The US wants to lead in AI.
It also wants to control who touches the most powerful models. Those two goals pull in opposite directions. The Mythos 5 case is a test run.
How many more such suspensions will come? How many more models will face sudden restrictions?
Anthropic is one company. Mythos 5 is one model. The pattern, however, is not unique.
Other frontier AI developers face similar scrutiny. The government is learning on the job.
So are the companies. The resolution this time was cooperation. Next time, the stakes could be higher.
The speed of the reversal shows that both sides understand what is at risk: security, leadership, and trust.
{“@context”: “https://schema.org”, “@type”: “NewsArticle”, “headline”: “US Commerce Secretary lifts ban on Anthropic AI for government work”, “description”: “Two weeks ago, the US government slammed the brakes on Anthropic’s most powerful cybersecurity AI. The order was blunt: bar foreign nationals from accessing Mythos…”, “datePublished”: “2026-06-27T07:07:31+00:00T09:30:00+00:00”, “author”: {“@type”: “Person”, “name”: “Noah King”}, “publisher”: {“@type”: “Organization”, “name”: “InfoPulse Today”, “url”: “https://newsfeed.infopulsetoday.com/”}}
