AI Model Claude Mythos Sparks Debate Over Its Capabilities and Release

The hype surrounding Anthropic’s Claude Mythos has drawn a lot of attention, urging organisations and leaders to try getting access to this ‘revolutionary’ AI model that’s apparently too dangerous to be released to the public. On the occasion, the Godfather of AI and Meta’s former Chief Scientist, Yann LeCun , has sharply criticised the growing alarm surrounding Mythos, calling the hysteria “BS from self-delusion.” LeCun made the comment on X (formerly Twitter) in response to claims that the powerful new AI can rapidly uncover thousands of high-severity vulnerabilities in every major operating system and web browser — capabilities that reportedly once took security researchers months to identify. There’s been a widespread concern in the tech and financial sectors after Anthropic chose not to release Mythos publicly due to its advanced cybersecurity prowess. Instead, the company has shared it selectively with partners, including Apple , Google, and Microsoft, under a $100 million defensive initiative called Project Glasswing. The AI model’s reported abilities triggered an emergency meeting involving Federal Reserve Chair Jerome Powell and Treasury Secretary Scott Bessent with CEOs of major US banks, especially as Wall Street scrambled to assess potential risks and defensive applications. Cybersecurity firms such as CrowdStrike , Cisco, and Palo Alto Networks have described Mythos as a genuine inflexion point. However, prominent AI voices are pushing back. LeCun argued that similar vulnerability analysis can likely be achieved with smaller, cheaper models. LeCun’s view is also echoed by others. Gary Marcus called the threat “overblown” and suggested “we were played,” while David Sacks pointed to Anthropic’s history of what he sees as scare tactics. Dave Kasten of Palisade Research added that Anthropic is “a little ahead, but not overwhelmingly ahead” and lacks a permanent technological moat. He also noted that OpenAI is believed to have comparable capabilities it also plans to restrict its AI model. Lately, Anthropic’s revenue has reportedly tripled to over $30 billion amid the buzz, and the company is eyeing potential IPOs this year, just like rival OpenAI . Critics argue that the restricted release of Mythos may serve greatly as a marketing and safety narrative rather than being a genuine protective measure. This may be helpful to the firm in the AI race against giants like Google and OpenAI, especially after the fallout with the US Department of Defense earlier this year. Anthropic has largely focused on enterprise solutions, leading rival firm OpenAI to reconsider its business strategy of aligning its goals away from the consumer-centric model it chose earlier.