Media: The NSA used Mythos from Anthropic - ForkLog: cryptocurrencies, AI, singularity, the future

# Media: NSA Used Mythos from Anthropic

The U.S. National Security Agency (NSA) uses the high-profile Mythos model from Anthropic, despite the startup’s conflict with the Pentagon. Axios reports this, citing sources.

Earlier, disagreements arose between Anthropic and the Pentagon because the company refused to allow the use of AI models for mass surveillance of U.S. citizens and the development of autonomous weapons. As a result, President Donald Trump instructed all federal agencies to completely stop using the company’s AI technologies within six months.

In April, Anthropic created a new model, Claude Mythos, but refused to release it to the public due to high security risks. During several weeks of testing, the neural network discovered thousands of zero-day vulnerabilities in major operating systems and web browsers.

Instead of a public release, the company launched Project Glasswing — an initiative involving AWS, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorgan Chase, Linux Foundation, Microsoft, Nvidia, Palo Alto Networks, and others to test the tool in secure conditions.

Axios found that the NSA is among the undisclosed recipients of the technology. It is unknown how the agency uses the AI startup’s technology, but others are applying it to scan their own environments for vulnerabilities.

“At present, the military is expanding the use of Anthropic tools while simultaneously asserting in court that their use threatens U.S. national security,” journalists noted.

Relations between Anthropic and authorities have probably warmed. On April 18, CEO Dario Amodei discussed with representatives of the administration the possibilities for joint work.

The meeting included White House Chief of Staff Suzy Wiles and Treasury Secretary Scott Bessent.

Recall that in April, Anthropic released a platform for launching complex and long-lasting agent tasks — Claude Managed Agents.