At Microsoft Ignite 2025, Stuart Clark, senior VP of product at KnowBe4, emphasized the need to adapt security strategies for both human users and AI agents. Organizations are now recognizing that AI, making decisions at machine speed, should be treated similarly to human staff, complete with identities and risk profiles. Clark asserts that understanding user behavior—highlighted by data indicating that users often click email links within just 13 seconds—can enhance security training and create a unified trust framework across various tools. He detailed how the transition from email-centric to agent-centric workflows necessitates a shift in security training, as attackers increasingly manipulate AI language models. By establishing risk scoring and guidelines for autonomous agents, organizations can bolster their security posture. With over 20 years in cybersecurity, Clark also leads the development of AI-native security solutions at KnowBe4, previously contributing on similar fronts at Microsoft and IBM.
Source link
