US NSA Reportedly Uses Anthropic’s Claude Mythos AI for Cyber Operations

By: Harsh Vardhan

|

Updated on: 05-Jun-2026 03:00 PM

5,463 views

Follow Us:

5,463 views

The US National Security Agency (NSA) is reportedly using Anthropic’s advanced AI model, Claude Mythos, for cyber operations. According to a Financial Times report, Anthropic is assisting the NSA in deploying Mythos, which has identified thousands of critical cybersecurity flaws. The model’s capabilities have raised concerns about its potential use in cyberattacks against foreign networks.

Key Highlights

• NSA reportedly uses Anthropic’s Claude Mythos AI for cyber operations targeting foreign networks.
• Anthropic sent Forward Deployed Engineers to NSA to help deploy and customize the AI model.
• Claude Mythos access to expand to 150 organizations in 15 countries including India.

Claude Mythos and NSA Collaboration

Claude Mythos is regarded as one of the most advanced AI models for cybersecurity. Anthropic claims the model is so powerful that it cannot be released to the public at this time. The NSA may be leveraging Mythos to plan and execute cyberattacks on networks in countries such as China and Iran. The AI model can help identify and exploit vulnerabilities, potentially granting the US access to sensitive information or foreign networks.

Anthropic has reportedly sent about six Forward Deployed Engineers (FDEs) to the NSA. These engineers assist clients in integrating and customizing AI models for specific needs. At the NSA, the FDEs are likely supporting the deployment and adaptation of Mythos for national security purposes, including offensive cyber operations.

One source close to Anthropic stated, “The best way to build a good defence is to build a good attack.” The source added that if the US does not use Mythos for offensive purposes, adversaries might do so instead. However, it remains unclear if Anthropic’s engineers are directly involved in the NSA’s use of Mythos. There is no official confirmation on the model’s exact role within the agency.

Public and Industry Reactions

Reports of the NSA’s use of Mythos have sparked debate online. Some users on X (formerly Twitter) expressed concerns that governments intend to use AI to exploit, rather than patch, vulnerabilities. Others noted that it is unsurprising for governments to use advanced AI for offensive cyber activities. Some users criticized Anthropic for working with the NSA after previously refusing to collaborate with the Pentagon, calling the company’s stance inconsistent.

Expansion of Claude Mythos Access

Claude Mythos was initially available only to a select group of US-based organizations following its April release as part of Project Glasswing. The model’s ability to detect and exploit software vulnerabilities has drawn attention from governments, financial institutions, and IT companies worldwide. Countries, including India, have requested broader access to the technology.

Anthropic recently announced plans to expand access to Claude Mythos to approximately 150 additional organizations across 15 countries, including India. The company has not disclosed which Indian organizations will receive access. The expansion comes amid ongoing concerns about the model’s security implications and calls for fair international access.

Background and Policy Tensions

Anthropic has experienced tensions with the US government in the past. Earlier this year, the Pentagon labeled the company a supply chain risk after disputes over the unrestricted use of AI. In February, President Donald Trump ordered all US federal agencies to stop using Anthropic’s technology, stating, “We don’t need it, we don’t want it, and will not do business with them again.”

Despite these challenges, Anthropic continues to develop and expand access to Claude Mythos, while governments and organizations worldwide monitor its impact on cybersecurity and digital infrastructure.