The growing integration of advanced AI models like Anthropic’s Mythos is intensifying concerns about cyber risks, according to Kara Sprague, CEO of HackerOne. While select organizations are using these models to uncover security vulnerabilities, regulators and corporate leaders are wary of unintended consequences. Sprague highlights that the same tools designed to strengthen defenses could also be exploited by malicious actors, potentially increasing overall exposure. The dual-use nature of such technologies complicates risk management strategies across industries.
In a discussion with Caroline Hyde and Ed Ludlow on “Bloomberg Tech,” Sprague emphasized that cybersecurity is no longer just about protecting systems—it now involves managing the risks introduced by rapidly evolving artificial intelligence. As companies adopt AI for threat detection, they must also consider how those same systems might be weaponized. The balance between innovation and safety remains fragile, requiring proactive oversight and collaboration between technologists, businesses, and policymakers.