# Enterprises Strengthen AI Agent Security

In an October 6, 2026 interview, Ninad Katkar, Cisco India & South Asia’s security leader, described governing AI agents as identifiable digital actors through least-privilege access, continuous checks on intent, observability, and rapid access revocation. The discussion identifies no government solicitation, contract, budget, or procurement contact, so it is a market signal rather than a specific public-sector opportunity.

- Contractors offering identity and access management, AI security, or agent-monitoring capabilities can use these themes to inform relevant product and service offerings.
- Procurement professionals evaluating enterprise or AI systems can consider whether requirements address agent identity, permitted actions, accessed data and tools, activity visibility, and immediate revocation of access.
- The interview reflects a security concern raised by an industry executive; it does not establish a government mandate or a procurement timeline.

**Jurisdictions:** international
**Industries:** Information Technology
**Topics:** Cybersecurity, Artificial Intelligence
**Published:** October 06, 2026

### Vendors
- Cisco ()

### Key Quotes
> Every agent should answer five questions. What is this agent all about? What is it allowed to do? Why and what is it trying to do? So, what data and tools is it accessing? And can we stop that access immediately?
> — Ninad Katkar, Leader, Security, Cisco India & South Asia

### Sources
- [Identity and intent become the new pillars of enterprise security](https://www.expresscomputer.in/interviews/in-the-agentic-era-identity-and-intent-become-the-new-pillars-of-enterprise-security/139472) - Express Computer