Topic starter
11/07/2026 1:53 pm
TL;DR: AI agents create a containment problem because runtime behaviour can outpace endpoint-based enforcement and human review, according to Elisity. Network enforcement has to match where agents actually move, or lateral paths and shadow connections stay open.
Editorial analysis by NHI Mgmt Group, based on content published by Elisity: “AI Agent Network Security: Why Microsegmentation Is the Missing Layer”.
By the numbers:
- 48% of security professionals rank agentic AI as the #1 attack vector.
- 88% of organizations reported AI agent security incidents in the past year.
- Only 14.4% of AI agents deploy with full security approval.
Key questions
Q: Why do AI agents create more risk than traditional automation?
A: AI agents create more risk because they can interpret context, choose actions, and invoke tools autonomously.
Q: What are the signs that shadow AI is creating governance and compliance gaps?
A: Common signs include limited visibility into who is using AI tools, no reliable audit trail for prompts and outputs, unclear data residency, and inconsistent handling of sensitive information.
Q: What breaks when AI agents are not contained at the network layer?
A: Without network-layer containment, AI agents can still traverse east-west paths, reach services they should not use, and spread damage far beyond the original compromise or shadow deployment.
Practitioner guidance
- Map every AI agent to a network identity Create an inventory of sanctioned and shadow AI agents, then assign each one a distinct identity record that links it to reachable services, protocols, and owners.
- Constrain network paths by least privilege Define allowed east-west and outbound communication paths for each agent, then deny all other network routes by default so lateral movement has no available path.
- Separate containment from endpoint trust Do not rely on endpoint tooling alone to contain AI agents.
Bottom line: AI agents create a containment problem because their runtime behaviour can outpace endpoint-based enforcement and human review.
Glossary terms: Identity-based Microsegmentation, Shadow AI, Data Plane, Authorised-Channel Abuse
Explore further
View Full Forum → | NHI Foundation Course → | Our Services → | Read the full analysis →
This topic was modified 13 hours ago by NHI Mgmt Group
11/07/2026 2:05 pm
Network enforcement becomes the primary containment layer when the actor is autonomous. AI agents do not wait for human-paced review cycles, and that changes which control matters most. A policy that depends on endpoint visibility or manual intervention is too slow once the agent can discover and traverse the environment at runtime. The implication is that security programmes need a control plane that can govern movement independently of the agent's own execution context.
A few things that frame the scale:
- Systems with least-privileged AI access had a 17% incident rate vs 76% for over-privileged systems. Organisations failing to scope AI access properly are 4.5x more likely to experience a security incident, according to the 2026 Infrastructure Identity Survey.
- 70% of organisations grant AI systems more access than they would give a human employee performing the exact same job, according to the 2026 Infrastructure Identity Survey.
A question worth separating out:
Q: Should organisations treat microsegmentation as part of identity governance for AI agents?
A: Yes. For agentic systems, segmentation is not just a network design choice. It is an identity control that limits blast radius by constraining where an agent can go after access is granted, which is exactly where creative misuse becomes costly.
👉 Read our full editorial: AI agent network security needs identity-based microsegmentation