GenAI Security: Defending Against Deepfakes and Automated Social Engineering
In the latest InfoQ session, QCon AI New York 2025 chair Wes Reisz sits down with Shuman Ghosemajumder (Google Trust & Safety founder and Reken CEO) for a no-holds-barred chat about how deepfakes and automated social engineering are supercharging cybercrime by eroding digital trust. Shuman lays out why “default trust” is a liability and how AI-driven imposters are mimicking human behavior at scale.
To fight back, he argues we need to instrument systems with behavioral telemetry and lean on game theory to spot and stop attacks that act like people but aren’t. It’s all about moving away from “trust first, verify later” and toward real-time “track, test, and thwart.”
Watch on YouTube
Top comments (0)