Anthropic recently introduced Mythos, its latest advanced AI model tailored for defensive cybersecurity tasks. The model’s extensive capabilities have raised concerns about potential threats to conventional software security, as the AI startup disclosed uncovering “thousands” of significant vulnerabilities across major operating systems and web browsers.
While some industry experts have questioned the validity of Anthropic’s claims regarding the power of their AI technology, even critics acknowledge that Mythos may signify a significant advancement in AI technology.
The model was launched by Anthropic through a controlled program named “Project Glasswing,” providing access to tech giants like Amazon, Microsoft, Nvidia, and Apple. Additionally, over 40 other organizations involved in critical software infrastructure maintenance were granted access.
Experts have warned that Mythos has the ability to detect and exploit previously unknown vulnerabilities at a faster rate than companies can address them. This advanced model’s coding and autonomous functions could expedite complex cyberattacks, especially in sectors like banking that heavily rely on intricate and interconnected legacy technology systems.
During its launch, Anthropic expressed concerns that if misused, Mythos’s capacity to identify software flaws at scale could pose severe risks to economies, public safety, and national security. Following the Mythos debut on April 7, U.S. software stocks experienced a decline on April 9, reigniting fears of AI advancements disrupting traditional businesses.
Regulators, including Bank of Canada Governor Tiff Macklem, have emphasized the need for global financial systems to address the risks posed by rapidly evolving AI models like Mythos. Discussions have taken place among financial sector representatives to assess the potential impact of Mythos on the financial system.
The White House engaged in discussions with Anthropic CEO Dario Amodei regarding Mythos, focusing on collaboration, cybersecurity, and balancing AI innovation with safety. Despite the Pentagon’s supply-chain risk designation on Anthropic, the U.S. government plans to make a version of Mythos accessible to major federal agencies.
Concerns surrounding Mythos have extended to the U.S., Britain, and Germany, with authorities and financial institutions closely monitoring the model’s implications. Notable figures, including David Sacks, a prominent Anthropic critic, have urged stakeholders to take the concerns seriously, emphasizing the increasing capabilities of AI models like Mythos in identifying and exploiting vulnerabilities.