Zero Trust for AI Agents: Why Anthropic and Zscaler See It as Essential

A proven cybersecurity model is getting new relevance in the era of autonomous AI. Recent Anthropic’s paper points directly to Zero Trust as the ultimate strategy for securing agentic workflows. In this video, Dhawal Sharma, EVP, AI Security and Strategic Initiatives, breaks down why traditional perimeters fall short when AI agents gain agency—accessing tools, executing tasks, and reasoning at machine speed. Watch to learn how Zero Trust principles extend to non-human identities, contain agent sprawl, and enforce continuous protection across the modern AI attack surface.