The White House Opposes Anthropic Expanding Mythos to 120 Companies: Wants to Keep It for Themselves Due to Insufficient Hash Rate

According to DataBeat Monitor, Anthropic has proposed to add about 70 companies and institutions to use Mythos, in addition to the existing 50, bringing the total to about 120. The White House refused citing security and computing power concerns. Mythos was released in early April, capable of detecting and exploiting vulnerabilities in a large amount of critical software, currently only available for testing by enterprises managing critical infrastructure, with no public release schedule.

The White House is concerned about Anthropic's computing power. With 70 more commercial users onboard, using Mythos for government purposes may be constrained. White House AI advisor David Sacks has publicly stated that the limited release of Mythos itself indicates that its computing power is inferior to that of its competitors. Anthropic has already signed computing power procurement agreements with Amazon, Google, and Broadcom, but the new capacity has not yet come online.

There is also no easing of tensions politically. The Trump administration criticized Anthropic for hiring several former Biden administration officials and is displeased with its connections to liberal organizations. One detail illustrating the level of trust is that former Anthropic researcher Collin Burns had been arranged to take up a position in the government's AI model assessment team. Once senior White House officials learned of this, they promptly replaced him, stating that they did not want someone from an AI company in a position that would involve engaging with major AI companies.