Artificial intelligence is increasing both the clockspeed of organizations and the cyber risks they face. Anthropic’s unreleased model, Claude Mythos, which uses advanced reasoning to find vulnerabilities that have escaped detection for decades, should be a wake-up call. How can companies take advantage of AI’s promise without falling victim to its risks? Earlier, we argued that, by working together in a synchronized approach to digital risk, business, IT, and security teams can more effectively anticipate and manage cyber risks. But organizations are still having trouble keeping up. With bad actors introducing AI agents into their arsenal, the stakes are even higher than before. To understand where current cybersecurity operating models break down, what high-performing organizations do differently, and what leaders need to change, we benchmarked several leading companies.

Full analysis : Companies that accelerate AI adoption without redesigning their security operating model are putting themselves at risk.