Pathlock CEO on identity security in the AI era

At a glance

What this is: NHI Mgmt Group's analysis, drawing on Pathlock's report: Agentic AI is changing how organisations think about identity security, governance and compliance because static entitlement models do not match autonomous runtime behaviour.

Why it matters: It matters because IAM, IGA and PAM teams need to decide which controls still work when software can act, choose tools and sequence actions without waiting for a human gate.

Context

Agentic AI changes the identity problem when software can make runtime decisions and act without waiting for human approval. That breaks assumptions baked into IAM, IGA and compliance programmes that were designed around stable users, predictable access requests and review cycles.

Pathlock's interview frames this as a governance issue as much as a security issue. For practitioners, the key question is not whether AI is present in the stack, but whether identity controls can still govern autonomous behaviour at the point of action.

This is best read as a category signal rather than a product announcement. The article points to the growing mismatch between traditional entitlement management and systems that can choose, combine and execute actions dynamically.

Key questions

Q: What breaks when access review models are applied to agentic AI?

A: Access review models break when the actor can obtain, use, and release privileges before the review cycle sees a stable state. Agentic AI compresses the decision window so tightly that the entitlement may never exist long enough to certify, challenge, or revoke in a meaningful way. Governance has to shift toward runtime validation and action-level attribution.

Q: Why do AI-driven attacks change identity governance requirements?

A: Because they compress attacker decision cycles and increase the volume of abuse that identity controls must evaluate. Traditional review cadences assume defenders have time to investigate each event. AI-assisted fraud and account takeover shorten that window, so governance must focus on telemetry quality, containment speed, and accountable automation.

Q: How should organisations govern agentic AI when it makes judgment calls, not just automated actions?

A: Organisations should govern the decisions agentic AI is permitted to make, not only the data it can access. That means defining escalation thresholds, preserving human decision rights for ambiguous cases, and logging the signals that explain why the agent chose a path. If the decision itself is unreviewable, policy drift can occur without a visible breach.

Q: What is the difference between AI automation and agentic AI from an identity perspective?

A: AI automation performs predefined tasks within narrow boundaries, while agentic AI can choose actions, call tools, and persist across sessions. From an identity perspective, agentic systems need stronger authentication, authorization, review, and offboarding because they behave more like active identities than fixed scripts.

Technical breakdown

Why static entitlements break down for agentic AI

Static entitlements assume the actor's purpose, scope and timing are known at provisioning time. Agentic systems can change those variables during execution, selecting tools and actions as the task unfolds. That creates a mismatch between pre-approved access and actual runtime behaviour, especially when the same identity can touch multiple data sources or services in one session. The governance problem is not just excess privilege. It is that the access decision may be obsolete before review can ever happen.

Practical implication: Treat runtime authorisation as a separate control plane from provisioning and review.

How governance changes when software can act on its own

Agentic AI turns identity from a static account record into an operational decision surface. Traditional IGA assumes access is granted to a known subject, then recertified later. Autonomous behaviour collapses that window, because tool choice and execution timing happen mid-session. That means governance has to account for who or what initiated the action, which policy boundary applied, and whether the system stayed within its intended scope. This is where human-paced review processes lose fidelity.

Practical implication: Map decision authority, tool scope and approval boundaries before agents are allowed to execute.

Why compliance teams need evidence at runtime

Compliance models work best when they can point to a durable entitlement, a control owner and an auditable approval chain. Agentic AI complicates all three because the meaningful event is not just access grant, but action selection and execution. In practice, audit evidence must move closer to the transaction itself, including the policy context that allowed the agent to act. Without that, compliance reporting can show that access existed while missing whether the access was appropriate for the moment it was used.

Practical implication: Shift evidence collection from periodic access review to action-level logging and policy traceability.

NHI Mgmt Group analysis

Agentic AI creates an identity governance assumption collapse, not just a new access risk. Static entitlement models were designed for actors whose intent, scope and timing are stable enough to be certified later. That assumption fails when the actor can decide and execute inside the session. The implication is that access governance must stop assuming a durable reviewable state and start treating runtime behaviour as the control boundary.

Agentic systems turn identity from a provisioning problem into a delegation problem. Once software can choose tools and sequence actions, the central question becomes who authorised the behaviour, not merely who received the account. That shifts the governance lens from entitlement ownership to execution authority. Practitioners should read this as a sign that traditional approval chains are no longer sufficient evidence of control.

The named concept here is runtime governance gap. It describes the space between pre-approved access and the actual action an agent performs in the moment. In human IAM, this gap is usually manageable because actions unfold through slower, observable workflows. In agentic AI, the gap can close too fast for retrospective review to matter.

Compliance frameworks will increasingly be judged by their ability to explain action, not just access. If an autonomous system can act within policy but outside human expectation, the control story has to account for both. That does not mean abandoning IGA or PAM. It means recognising that their evidence model must evolve from static permission checks to runtime accountability.

Pathlock's interview signals a broader market shift toward execution-aware identity governance. The industry is moving beyond the question of who can sign in and toward what the identity is allowed to do once it is already active. That matters for NHI, human IAM and agentic AI alike, because all three now meet at the point where access becomes action.

From our research library:

• 88% of organisations have embedded AI agents in their workflows, according to KPMG's 2026 report.
• Read next: Agentic AI Identity Guide

What this signals

Runtime governance gap: access controls built for periodic review lose precision when software can act before the review cycle ever reaches the session. The operational focus shifts from entitlement count to decision timing, which is why agentic AI governance cannot be handled as a simple extension of human IAM.

The practical boundary for security teams is no longer the login event but the action boundary. If an agent can pick tools, call systems and complete work without a human checkpoint, then governance has to observe behaviour as it happens, not after the fact.

For practitioners

• Map runtime decision points Identify where an AI system can choose tools, data sources or actions after authentication, and separate those points from ordinary access approval flows.
• Define execution boundaries Document which actions remain human-approved, which can be delegated, and which require a policy gate before an agent can proceed.
• Instrument action-level logging Capture the action taken, the policy context and the data touched so audits can reconstruct what the system actually did, not just what it was allowed to do.
• Rework access reviews for agents Stop treating periodic recertification as sufficient evidence for autonomous systems and validate whether the underlying behaviour still matches the original approval intent.

Key takeaways

• Agentic AI exposes a mismatch between static entitlement models and software that can make decisions during execution.
• The main governance issue is not just access volume, but whether review-based controls can still prove appropriateness at runtime.
• Practitioners need execution-aware controls that separate permission to operate from permission to act.

Key terms

• Agentic AI: Autonomous AI systems capable of planning, deciding, and taking actions, including calling APIs, writing code, and orchestrating other agents, with minimal human oversight. Agentic AI introduces new NHI risks as agents must authenticate to external services.
• Runtime Governance: Runtime governance is the set of controls that verify what a system or agent is actually doing after deployment. It combines monitoring, authorization checks, and access validation so teams can detect drift, misuse, or excessive privilege in motion rather than assuming build-time policy still holds.
• Execution boundary: The point at which an authorised task turns into a real system change, such as writing data, deleting records, spending money, or invoking a downstream tool. In AI governance, controlling the execution boundary matters more than simply approving access, because harm occurs when actions are allowed to complete unchecked.
• Delegated decision authority: Delegated decision authority is the transfer of a bounded operational decision from a human to a system. For AI security operations, it marks the point where automation is no longer just executing rules, but influencing which incidents matter and what response should happen next.

Deepen your knowledge

NHI governance, agentic AI identity, and machine identity lifecycle are core topics in our NHI Foundation Level course, the industry's only accredited NHI security programme. If you are building or maturing an IAM programme, it is worth exploring.