Fears that Anthropic’s Mythos model would trigger a surge in malicious cyberattacks have been labeled as exaggerated by cybersecurity professionals one month after its April release. While Anthropic warned that the model had identified thousands of vulnerabilities across major operating systems and browsers, security experts argue that access to a model of this caliber does not immediately enable new hacking capabilities for bad actors. Isaac Evans, CEO of the security firm Semgrep, noted a significant communication gap between technical practitioners and policymakers, stating that the model's technical advances do not necessarily translate to a simplified exploitation process in the field.

Inside the industry, experts familiar with early testing of Mythos confirmed that it improves vulnerability discovery, but emphasize that the primary challenge remains the validation, prioritization, and remediation of these flaws without system disruption. Anthony Grieco, an executive at Cisco, explained that Mythos allows for faster code scanning and lowers the rate of false positives, which assists defenders in addressing the most critical risks. To effectively utilize these tools, however, organizations require substantial computing power and a rigorous harness—a secure environment with specific limitations for running language models.

Anthropic’s collaborative Project Glasswing, which invited select firms to test security defenses, helped raise public and governmental awareness about the model. This initiative drew attention from the White House, which is currently evaluating rules for releasing new AI models, even as the Pentagon identified Anthropic as a supply-chain risk. Industry experts stress that the ability to find vulnerabilities with AI is not a new development, and most real-world threats still originate from adversaries who operate without AI assistance. Current limitations, including the substantial computing infrastructure required to run Mythos, present temporary barriers to entry that are expected to diminish as technology evolves, according to Nick Adam of State Street.

Fears that Anthropic’s Mythos model would trigger a surge in malicious cyberattacks have been labeled as exaggerated by cybersecurity professionals one month after its April release. While Anthropic warned that the model had identified thousands of vulnerabilities across major operating systems and browsers, security experts argue that access to a model of this caliber does not immediately enable new hacking capabilities for bad actors. Isaac Evans, CEO of the security firm Semgrep, noted a significant communication gap between technical practitioners and policymakers, stating that the model's technical advances do not necessarily translate to a simplified exploitation process in the field.

Inside the industry, experts familiar with early testing of Mythos confirmed that it improves vulnerability discovery, but emphasize that the primary challenge remains the validation, prioritization, and remediation of these flaws without system disruption. Anthony Grieco, an executive at Cisco, explained that Mythos allows for faster code scanning and lowers the rate of false positives, which assists defenders in addressing the most critical risks. To effectively utilize these tools, however, organizations require substantial computing power and a rigorous harness—a secure environment with specific limitations for running language models.

Anthropic’s collaborative Project Glasswing, which invited select firms to test security defenses, helped raise public and governmental awareness about the model. This initiative drew attention from the White House, which is currently evaluating rules for releasing new AI models, even as the Pentagon identified Anthropic as a supply-chain risk. Industry experts stress that the ability to find vulnerabilities with AI is not a new development, and most real-world threats still originate from adversaries who operate without AI assistance. Current limitations, including the substantial computing infrastructure required to run Mythos, present temporary barriers to entry that are expected to diminish as technology evolves, according to Nick Adam of State Street.