Despite blacklist, NSA is reportedly using Anthropic’s Mythos: Report

The US National Security Agency (NSA) is reportedly using Anthropic’s new cybersecurity Artificial Intelligence (AI) model, Claude Mythos, despite the government’s ban and the Department of Defence “supply-chain risk” label. According to an Axios report, the NSA is among the 40 listed companies that have access to the Mythos. However, Anthropic did not reveal the names of the companies during its preview release.

NSA expands Anthropic’s Mythos use: What we know

If the claims are accurate, then it seems like the government is prioritising its cybersecurity needs over its ongoing conflict with Anthropic. Previously, the Pentagon urged government officials to stop using Anthropic’s tool. The government and the Trump administration are also running a legal battle against the label and broader ban on the company’s AI technologies.

Also read: Anthropic’s Claude Opus 4.7 is here: Check improvements, safety features, and what’s new

Despite the ongoing feud, the military is broadening the use of Anthropic's tools and technologies like Claude Mythos. However, it’s not clear exactly how the NSA is using the model. As per the release, the model could be used to scan systems for security weaknesses and identify potential cyber vulnerabilities.

The Mythros preview currently has restricted access to Mythos, and only about 40 organisations have access to it due to its capabilities being potentially risky. Anthropic released 12 organisations having access, and now, the NSA is believed to be among the unnamed users. In addition, the UK also has access via the AI Security Institute.

Reportedly, Anthropic CEO Dario Amodei on April 17 also met Susie Wiles, White House chief and Treasury Secretary Scott Bessent to discuss the government's use of Mythos, its security plans, and future plans. It is suggested that the meeting could map out how other government departments, apart from the Pentagon, can use its AI model.

Also read: Anthropic’s latest AI model ‘Mythos’ triggers urgent risk review by UK regulators: Report

Is Anthropic’s Mythos safe?

While Mythos is said to deliver advanced capabilities for cyber defence, concerns remain about its potential misuse if it falls into the wrong hands. Given its powerful features, access has been limited to select users, as the company continues to evaluate safety measures and carefully plan any broader public release.