The White House is reportedly moving to allow federal agencies to use Anthropic Claude Mythos, the company’s powerful new frontier AI model.
Gregory Barbaccia, the U.S. CIO, told Cabinet officials
in a memo sent Tuesday that the Office of Management and Budget was setting up protections to allow federal agencies to begin using a version of Mythos.
“We're working closely with model providers, other industry partners, and the intelligence community to ensure the appropriate guardrails and safeguards are in place before potentially releasing a modified version of the model to agencies," he wrote.
He did not specify when the model would become available.
It’s difficult to describe just how much Mythos has captured the imaginations of those in tech. The AI model—which is unreleased and available only to a restricted group—has gained
much attention in recent weeks for its potency.
As part of so-called
Project Glasswing, Mythos has become a sort of cybersecurity silver bullet to unearth high-severity vulnerabilities in the most critical software, such as what underpins the nation’s banking system.
Concerns abound, then, that Mythos falls into the wrong hands—only for corporate chiefs and government officials to find themselves on the receiving end of its powers.
—AN
