Earlier this year, Anthropic unveiled a preview version of Mythos, its upcoming AI model that it claimed was simply too dangerous to make available to the wider public.
As part of an early access program, the company allowed around 50 businesses and organizations to test it out, saying it was giving them a fighting chance to get ahead of a potential cybersecurity crisis in the making. According to Anthropic’s own researchers, Mythos is powerful enough to identify and exploit sweeping cybersecurity vulnerabilities, potentially giving hackers a major advantage.
In fact, the company admitted last week that it was investigating a potential unauthorized access to its Mythos model.
The alleged risks appear to have spooked even the White House. As the Wall Street Journal reports, Trump administration officials said they oppose Anthropic’s plan to expand access to Mythos to roughly 70 additional companies and organizations, citing security concerns.
One source told the newspaper that Mythos would suck up too many computing resources, which could hamper the US government’s ability to use it effectively, a claim Anthropic has since denied.
It’s a strange new development in Anthropic’s shaky relationship with the Department of Defense. The two have been caught in a surprisingly public feud. In late February, CEO Dario Amodei refused to allow the military to use its AI models for autonomous weaponry and the mass surveillance of US citizens, an insistence that greatly angered defense secretary Pete Hegseth and president Donald Trump.
Bafflingly, Anthropic is technically still listed as a supply chain risk today, a label White House officials slapped on the AI company seemingly in retaliation, forcing any defense contractors to sever their relationships. That’s despite the US military heavily relying on the AI models during its bombing of Iran.
However, department of defense CTO Emil Michael insisted during a CNBC interview today that Mythos was a “separate national security moment.”
In other words, the White House appears to both want unfettered access to Mythos, while also labeling its developer as a major threat to national security.
Meanwhile, an Anthropic spokesperson told the WSJ that the company is having productive conversations with the government about rolling out access to Mythos to more companies and organizations.
For now, we’ll have to wait and see if Mythos is all it’s cracked up to be — and whether it does warrant this unprecedented level of caution.
“A growing number of people are wondering if Anthropic is the AI industry’s ‘boy who cried wolf,’” White House AI advisor David Sacks tweeted last month. “If Mythos-related threats don’t materialize, the company will have a serious credibility problem.”
More on Anthropic: Top Security Experts Alarmed by Power of Anthropic’s New Hacker AI
The post The White House Suddenly Seems Pretty Terrified of Anthropic appeared first on Futurism.
