On #DigitalDialogues, we got an exclusive interaction with Amit Datar, VP – Product Management, Netskope, we explore the emerging AI security risks organisations must prepare for in 2026, from agent-led decision pathways to the expanding attack surface introduced by Model Context Protocol (MCP).
Generative AI is no longer operating in isolation, AI agents are increasingly interacting with enterprise systems, tools, and data through emerging protocols such as the Model Context Protocol (MCP). While this unlocks powerful automation and decision-making capabilities, it also introduces new security risks tied to machine-led access and context sharing.
In this interview, Amit Datar from Netskope outlines key cybersecurity trends for 2026, focusing on:
- AI-driven attack surfaces beyond GenAI models
- Security guardrails required for agentic AI deployments
- What MCP is and the risks it introduces
- Applying Zero Trust principles to LLM ecosystems
- Strategies to secure MCP-enabled environments through provenance enforcement, layered isolation, permission controls, and risk-based access
In an AI-native enterprise, trust will depend not only on identity – but on context.
