White House wants US agencies to use Claude Mythos despite fallout with Anthropic

The US government may be planning to give Anthropic a second chance. After a fallout over unrestricted AI use for the Pentagon, the White House is reportedly considering using Claude Mythos, Anthropic's most powerful AI system, across US agencies in efforts to bolster cybersecurity.

In Short

• White House is considering using Claude Mythos for US agencies
• Claude Mythos is believed to be most powerful model for cybersecurity
• Anthropic is designated as a supply chain risk after Pentagon fallout

Earlier this year, Anthropic found itself at odds with the US government after it refused unrestricted use of its AI systems for the military. The US Department of Defense swiftly labelled the company a supply chain risk, and decided to phase out its systems in favour of OpenAI. However, it seems that Anthropic has found a way to get back into the White House with Claude Mythos.

As per a report from Bloomberg, the White House is considering allowing US agencies to use Claude Mythos, an AI model so powerful that Anthropic decided not to release it to the general public. This comes at a time when Anthropic is fighting the supply chain risk designation in court.

Do note that in February this year, US President Donald Trump ordered all federal agencies to stop using Anthropic’s products. He stated, “We don’t need it, we don’t want it, and will not do business with them again!”

Why does the White House want Claude Mythos

Claude Mythos is Anthropic’s most powerful AI model yet. The company claims that the AI can expose cybersecurity vulnerabilities like never before. This has reportedly raised concerns amongst US government officials who want to use the model to bolster cyberdefenses.

As per the report, Gregory Barbaccia, federal chief information officer of the White House Office of Management and Budget (OMB), sent an email to US Cabinet departments earlier this week. Barbaccia told officials that OMB was setting up a way for agencies to use Mythos.

However, the email did not confirm which agencies will likely get access to Claude Mythos. Though it states that agencies can expect more information “in the coming weeks.”

It is believed that the US government wants to use Claude Mythos to identify potential software flaws across its own systems. According to a report from the Financial Times, “several chief information officers” across the US government have asked for access to Mythos.

What about the Pentagon fallout?

It is expected that if the White House considers Claude Mythos as a necessary tool for agencies, it may override the supply chain risk tag given to Anthropic by US Defense Secretary Pete Hegseth.

Anthropic is fighting this designation in US courts. The company claims that the designation could impact billions of dollars of revenue.

The US military is believed to still be using Anthropic’s AI models during its operations in Iran. Though it is expected to phase them out within a few months as part of a transition phase.

A US official told Financial Times, that the White House “continues to proactively engage across government and industry to protect the United States and Americans”

Claude Mythos access remains limited

Earlier this month, Anthropic announced Project Glasswing. This project gives access to Claude Mythos to a select group of companies, including Amazon, Microsoft and Google, as a way to fix potential software flaws flagged by the AI.

Prior to its limited release, reports indicate that Anthropic briefed senior US government officials on Mythos, highlighting both its offensive and defensive applications. The briefings included input from the Cybersecurity and Infrastructure Security Agency and the Center for AI Standards and Innovation.

On the day of Claude Mythos’ release, US Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell held a meeting with Wall Street leaders, urging them to use Mythos to identify potential flaws in their own systems.

- Ends