Deepfake Detection, AI Agents, and Continuous Identity
Vijay Balasubramaniyan · CEO and Co-Founder · Pindrop
A voice, a video interview, or an AI agent can appear legitimate without being trustworthy. Vijay Balasubramaniyan, CEO and Co-Founder of Pindrop, explains how deepfakes, automated fraud, and fake job candidates are changing the way companies verify identity.
Pindrop began by securing phone calls, long before voice interfaces and generative AI made the question of who is on the other end feel urgent. Vijay describes how the company spent 15 years studying human behavior and how quickly voice and video cloning tools have improved. A few seconds of audio can now be enough to clone a voice.
The conversation explores good and malicious AI agents. Vijay describes fraudsters using language models and text-to-speech systems to automate account attacks, while legitimate agents handle tasks such as employment checks and healthcare administration. As adoption grows, businesses must determine whether an agent is authorized, harmful, or has changed behavior, and who is responsible when an agent makes a damaging decision.
Hiring is another major focus. Vijay discusses fake applicants, identity changes between interview rounds, and why recruiters increasingly have to think like security teams. He makes the case for continuous identity: verifying the same person throughout interviews, work, and important virtual interactions. Stronger ongoing signals may also let organizations make security less visible for legitimate users.