Meta's AI Pioneer Yann LeCun Dismisses Anthropic's Claude Mythos as Overblown Theatrics
Yann LeCun, Meta's former chief AI scientist and one of the three researchers credited with pioneering modern deep learning, has dismissed the panic surrounding Anthropic's Claude Mythos Preview as overblown theatrics. In a blunt post on X, LeCun wrote, "Mythos drama = BS from self-delusion," responding to claims from AI security company Aisle that smaller, cheaper models could replicate much of the vulnerability analysis showcased by Anthropic.
LeCun's Comments Follow High-Profile Claims and Emergency Meetings
LeCun's comments came just days after Anthropic announced that its Mythos Preview had identified thousands of high-severity vulnerabilities across every major operating system and web browser. These claims were serious enough to prompt Federal Reserve Chair Jerome Powell and Treasury Secretary Scott Bessent to convene an emergency meeting with CEOs of America's largest banks. The model, which Anthropic has refused to release publicly, is being shared exclusively with a select group of companies including Apple, Google, and Microsoft under a $100 million initiative called Project Glasswing.
LeCun Isn't the Only Voice Calling It Hype
LeCun's skepticism is not isolated. AI researcher and frequent industry critic Gary Marcus wrote on Substack that the Mythos threat was "overblown" and added, "to a certain degree, I feel that we were played." Marcus argued the model appears to be only "incrementally better" than its predecessors rather than representing a true breakthrough. David Sacks, the former White House AI czar, acknowledged the cybersecurity risks but noted 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 lacks a permanent moat, pointing out that OpenAI reportedly has a model with similar capabilities it also plans to restrict.
Timing Raises Questions Amid Financial Growth
The timing of Anthropic's announcement has not helped its case. The company's projected annual revenue has tripled to over $30 billion this year, and both Anthropic and OpenAI are reportedly exploring initial public offerings. A model deemed too dangerous for public release, available only to the world's largest corporations, reads as easily as a safety decision as it does a strategic sales pitch, fueling further debate about the company's motivations.
But Companies Using Mythos Tell a Different Story
Not everyone agrees with LeCun's dismissal. The organizations that have been hands-on with Mythos Preview for weeks—including Cisco, CrowdStrike, and Palo Alto Networks—are treating it as a genuine inflection point in cybersecurity. CrowdStrike's CTO Elia Zaitsev stated, "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 merely Anthropic's marketing department, lending credibility to their assessments of the model's impact.
The debate highlights a growing divide in the AI community between those who see Mythos as a marketing-driven exaggeration and those who view it as a transformative tool in cybersecurity. As Anthropic continues to restrict access and hype its capabilities, the controversy is likely to intensify, with implications for how AI advancements are communicated and regulated in sensitive fields like national security and finance.
