Topic starter
06/06/2026 2:35 am
TL;DR: AI is increasing both defensive and offensive activity, and Imprivata argues that traditional IAM cannot fully govern autonomous AI agents because attribution, dynamic least privilege, continuous monitoring, and human oversight must all work together, according to Imprivata. The old model assumes identities stay within human-paced control loops; autonomous agents break that assumption by acting continuously across systems.
Editorial analysis by NHI Mgmt Group, based on content published by Imprivata: “Imprivata helps organizations secure critical access and strengthen continuous monitoring for the AI era”.
Key questions
Q: How should security teams govern AI agents that inherit authority from other identities?
A: Security teams should govern AI agents by tracking identity lineage, not just credentials.
Q: Why do traditional IAM and DLP controls fall short for agentic AI?
A: Traditional IAM and DLP controls assume risk can be judged at a point in time from one request or one response.
Q: What are the signs that an AI agent has more access than it needs?
A: Common signs include access to systems outside the declared task, frequent permission exceptions, policies that are broader than the observed workflow, and usage logs showing actions the agent did not need to complete its job.
Practitioner guidance
- Define agent identities explicitly Create a distinct identity model for each autonomous AI agent so actions, owners, and policy boundaries are attributable rather than buried inside generic service accounts.
- Apply runtime least privilege Scope agent access by task, context, and risk at execution time, not only at provisioning time, so access can shrink when the environment or action changes.
- Instrument continuous behaviour monitoring Collect telemetry that shows what the agent attempted, which systems it touched, and when the behaviour deviated from expected patterns.
Bottom line: Autonomous AI agents expose the limits of traditional IAM because static roles and periodic reviews cannot keep pace with continuous, cross-system behaviour.
Explore further
View Full Forum → | NHI Foundation Course → | Our Services → | Read the full analysis →
06/06/2026 4:20 am
Traditional IAM assumes access is stable long enough to be reviewed, and that assumption fails when the actor is autonomous. Agentic AI does not wait for human-paced certification cycles or predictable session boundaries. It can act continuously across systems, which means governance has to account for runtime behaviour rather than only provisioned entitlements. The implication is that identity programmes need to rethink what review, approval, and attribution mean when the actor can change state faster than the control loop can observe it.
A few things that frame the scale:
- 53% of security leaders expect AI to run major portions of their infrastructure autonomously within the next three years, according to the 2026 Infrastructure Identity Survey.
- Only 13% of organisations feel extremely prepared for the reality of agentic AI despite the majority racing toward autonomous adoption, according to the 2026 Infrastructure Identity Survey.
A question worth separating out:
Q: How do teams keep autonomous AI actions accountable under zero trust?
A: Teams keep autonomous AI actions accountable by tying each action to a named identity, an approved context, and an audit trail that can be reviewed after the event. Zero trust does not remove the need for ownership; it increases the need for continuous verification and clear decision records.
👉 Read our full editorial: Agentic AI identity management exposes the limits of traditional IAM