White House wants Claude's Mythos model despite Pentagon-Anthropic feud: 'Don't give a f*** about War's position'

Claude's powerful new AI model, Mythos Preview, has caught the attention of White House which wants to deploy it across parts of the US federal government even as the Pentagon remains locked in a bitter fued with Dario Amodei's company.

Claude had earlier refused to allow its AI to be used for fully autonomous warfare or mass surveillance of Americans, leading to a stand-off with the Trump administration.

However, the White House is now engaging with the company to use its new model, which was stopped from being rolled out to public due to its powerful capabilities.

Despite tensions, officials acknowledge that the company's technology could bolster national security -- or pose risks if misused, according to a report in Axios.

Anthropic is currently limiting access to Mythos Preview to a small group of organisations to study its advanced cyber capabilities and strengthen safeguards. Some US government agencies are now seeking to be part of that group.

Mythos has found "thousands" of major vulnerabilities in operating systems, web browsers and other software. Its capabilities to code at a high level have given it a potentially unprecedented ability to identify cybersecurity vulnerabilities and devise ways to exploit them, Reuters quoted experts as saying.

According to the Axios report, the Office of Management and Budget has already reached out internally, saying it is reviewing whether agencies can use the model.

Two sources told Axios that discussions with Anthropic are underway and access could be granted within weeks.

Pentagon stand-off

Earlier, the Pentagon barred Anthropic from its contracts, branding it a "supply chain risk" and directing vendors to remove its software from defence workflows.

However, the company can continue working with non-military arms of the government while litigation continues.

"There's progress with the White House. There's not progress with [the Department of] War," an administration official told Axios.

Another official defended the evaluation process but criticised Anthropic's approach. "The government 'has a responsibility to evaluate every model to see where the frontier of tech is,'" the official was quoted as saying by Axios. The official added that the company was using "fear tactics" about Mythos' hacking potential and is "succeeding".

At the same time, internal divisions remain stark. "All the intel agencies use Anthropic. Every agency except War (Pentagon) wants to. That's because Anthropic doesn't want to kill people and War's position is 'don't tell us what the f*** to do.' But if you're the Department of Energy, you don't give a f*** about that. You're worried about the Chinese attacking the energy grid. So you want Anthropic," another official told Axios.

Anthropic has maintained that its models will not be used for mass surveillance or fully autonomous weapons.

Meanwhile, civilian agencies, including those overseeing energy and finance, see potential benefits of Claude's AI Model. Access to Mythos could help identify vulnerabilities in critical infrastructure and improve preparedness against cyber threats.

One Defence official previously told Axios that talks continued because "these guys are that good."