TL;DR: 54% of organisations have already had an AI-agent security incident, Willow says in survey evidence that points to overprivileged, underlogged fleets. AI security posture management has to move governance into discovery, issuance, and runtime if agents are to stay auditable.
Editorial analysis by NHI Mgmt Group, based on content published by Willow: “AI Security Posture Management: What It Is and How It Applies to AI Agent Fleets”.
By the numbers:
- 54% of organizations have already experienced a security incident tied to AI agents.
Key questions
Q: Why do AI agent fleets increase security risk so quickly?
A: A fleet multiplies the problem because each agent may connect to different tools, carry different credentials, and operate without a clear human owner.
Q: How do security teams know if AI governance is working?
A: Look for evidence that access decisions are reviewable, permissions are revocable, and exceptions are not becoming permanent.
Q: What breaks when AI agents are treated like standard human users?
A: You lose visibility into effective permissions, expected behaviour, and real blast radius.
Practitioner guidance
- Build a complete agent inventory Track every sanctioned and unsanctioned AI agent, including tool connections, owners, and the credentials each one uses.
- Assign governed identities to agents Bind each agent to an identity in the organisation's identity provider so that provisioning, revocation, and accountability are managed centrally.
- Reduce permissions to the action level Replace broad app-level grants with task-specific permissions so an agent can perform only the exact operations its workflow requires.
Bottom line: AI security posture management is the control discipline that closes the gap between agent sprawl and identity governance.
Glossary terms: AI Security Posture Management, Shadow AI, Action-Level Permissioning, Runtime Guardrail
Explore further
View Full Forum → | NHI Foundation Course → | Our Services → | Read the full analysis →
AI security posture management is becoming the missing governance layer for agent fleets. Traditional IAM assumes identities are relatively stable, request access in predictable ways, and can be reviewed after use. AI agents break that model because they can appear quickly, operate continuously, and touch many systems before a human ever looks at the trail. The practical implication is that identity governance must move from periodic review to continuous posture control.
A few things that frame the scale:
- 28% of secrets incidents now originate outside code repositories, in Slack, Jira, and Confluence, and are 13% more likely to be categorised as critical than code-based leaks, according to the State of Secrets Sprawl 2026.
- Only 23% of IT leaders were very confident in their organisation's ability to manage security and governance for GenAI deployments, according to a 2025 Gartner survey of 360 IT leaders.
A question worth separating out:
Q: What should teams do when employees deploy shadow AI without approval?
A: Identify the agents first, then bring them into a governed identity framework with scoped access and audit logging. The goal is not always immediate shutdown. The practical decision is whether the agent can be made accountable and constrained quickly enough to reduce exposure before it becomes embedded in operations.
👉 Read our full editorial: AI security posture management exposes governance gaps in agent fleets