Artificial intelligence significantly intensifies cyber risk by enabling attacks that are highly automated, adaptive and scalable, such as generative systems that create tailored phishing at speed and drive a sharp increase in successful intrusions across sectors. As AI-automated attack chains, shadow AI and growing regulatory pressure make flawless prevention unrealistic, resilience becomes essential and is understood as the ability to “absorb, adapt and quickly recover from incidents,” shifting focus from stopping every breach to limiting impact and accelerating recovery.
Building this resilience starts with modernizing and securing AI foundations through security-by-design across data, models, applications and identity, addressing threats like prompt injection and data poisoning, governing AI in line with new rules, and replacing legacy tools with AI-ready platforms that support analytics, predictive threat modelling and automated remediation, a step many organizations still lack according to recent industry reports. With these foundations, AI can be used to automate detection and response, reduce false positives, enhance security operations, strengthen identity and access management, and speed tasks such as vulnerability scanning and cyber intelligence reporting, provided that change management, clear guardrails for autonomous actions and workforce training keep humans in the loop.
In a more advanced stage, cybersecurity is reinvented through agent-first operations where autonomous AI agents act as active defenders, managing identities, monitoring attack surfaces, reviewing contracts for missing security controls and conducting agent-led penetration testing, which allows threats to be anticipated and neutralized before they fully materialize. Overall, “autonomous agents are changing the rules of cybersecurity and inaction carries real risk,” urging organizations of all sizes to invest now in foundational AI security and skills to ensure they develop the intelligent resilience needed to stay ahead of AI-driven threats rather than being overtaken by them.
Reference
Kendzior, D., & Hosner, C. (2026, February 9). How can we build intelligent resilience against cyber threats in the age of AI. World Economic Forum. https://www.weforum.org/stories/2026/02/how-can-we-build-intelligent-resilience-against-cyber-threats-in-the-age-of-ai/