Kamal Ghorui
Yann LeCun—Meta’s former chief AI scientist and one of the three researchers credited with pioneering modern deep learning—has dismissed the panic around Anthropic’s Claude Mythos Preview as overblown theatrics. “Mythos drama = BS from self-delusion,” LeCun wrote on X, responding to a post from AI security company Aisle that claimed smaller, cheaper models could replicate much of the same vulnerability analysis Anthropic showcased in its announcement this week.LeCun’s comments came days after Anthropic said Mythos Preview had found thousands of high-severity vulnerabilities across every major operating system and web browser—claims serious enough to prompt Fed Chair Jerome Powell and Treasury Secretary Scott Bessent to convene an emergency meeting with CEOs of America’s biggest banks. The model, which Anthropic has refused to release publicly, is being shared only with a select group of companies including Apple, Google, and Microsoft under a $100 million initiative called Project Glasswing.
LeCun isn’t the only one calling it hype
LeCun’s scepticism isn’t isolated. Gary Marcus, the AI researcher and frequent industry critic, wrote on Substack that the Mythos threat was “overblown” and that “to a certain degree, I feel that we were played.” Marcus argued the model appears to be “incrementally better” than its predecessors rather than a breakthrough. David Sacks, the former White House AI czar, acknowledged the cybersecurity risks but added that “it’s hard to ignore that Anthropic has a history of scare tactics.” Dave Kasten from Palisade Research told CNBC that Anthropic is “a little ahead, but not overwhelmingly ahead” and doesn’t have a permanent moat—noting that OpenAI reportedly has a model with similar capabilities it also plans to restrict.The timing hasn’t helped Anthropic’s case. The company’s projected annual revenue has tripled to over $30 billion this year, and both Anthropic and OpenAI are reportedly exploring IPOs. A model too dangerous to release, available only to the world’s largest corporations—that reads as easily as a safety decision as it does a sales pitch.
But the companies actually using Mythos tell a different story
Not everyone agrees with LeCun. The organisations that have been hands-on with Mythos Preview for weeks—Cisco, CrowdStrike, Palo Alto Networks—are treating it as a genuine inflection point. CrowdStrike’s CTO Elia Zaitsev said “what once took months now happens in minutes with AI.” Cisco’s chief security officer called it a threshold moment with “no going back.” These are cybersecurity companies with their own reputations on the line—not Anthropic’s marketing department.
