Extended access management for AI agents and trusted devices

At a glance

What this is: This is 1Password’s analysis of how Extended Access Management aims to secure access across people, apps, devices, and AI agent workflows.

Why it matters: It matters because IAM teams now have to govern human sign-in, SaaS sprawl, device posture, and non-human access in one operating model instead of treating them as separate problems.

👉 Read 1Password's article on Extended Access Management for people, apps, devices, and AI

Context

Extended access management is the idea that access control has to follow people, devices, applications, and AI-driven workflows at the same time. The article argues that legacy IAM and IGA leave gaps in application discovery, credential handling, and governance when work is distributed and fast moving.

For identity teams, the question is no longer whether a user can sign in. It is whether access is only granted from trusted devices, whether unmanaged SaaS is visible, and whether agentic workflows are governed with the same discipline as human access. That is the real operating gap the article points to.

The article’s starting position is typical of modern enterprises: speed, collaboration, and distributed access are normal, while identity controls are still catching up.

Key questions

Q: How should security teams govern AI agent access alongside human IAM?

A: Security teams should keep human sign-in, machine access, and AI agent workflows in separate governance views even when they touch the same application. That makes ownership, entitlement review, and revocation clearer. The goal is not to treat agents like people, but to ensure their access is explicit, bounded, and auditable.

Q: Why do unmanaged SaaS apps create identity governance risk?

A: Unmanaged SaaS creates risk because access can exist outside normal onboarding, offboarding, and review processes. If identity teams cannot see the application, they cannot verify who has access, whether credentials are still active, or whether the app should be deprovisioned. Visibility is the prerequisite for lifecycle control.

Q: When should organisations require device trust before sign-in?

A: Organisations should require device trust for sensitive systems, remote work, and any environment where unmanaged devices could expose credentials or data. Device trust is most valuable when access decisions need to account for endpoint health, not just user authentication. It reduces the chance that valid credentials are used from unsafe hardware.

Q: Who is accountable when shared credentials are used across teams?

A: Accountability should sit with the service owner or platform owner, not with the people who happen to use the credential day to day. Shared access needs a named owner, logging, and a revocation path. Without that, offboarding and review become guesswork instead of a controlled process.

Technical breakdown

Why trusted device enforcement matters for access control

Device trust adds a conditional layer to access decisions by checking whether the laptop, tablet, or phone meets policy before sign-in is allowed. In practice, this turns device posture into part of the authorisation decision rather than a separate endpoint problem. The article’s model assumes that access should be gated by whether the endpoint is known, compliant, and trusted, especially when work happens across remote and mobile environments. That matters because sign-in alone does not prove that the device is safe enough to reach sensitive systems.

Practical implication: use device posture as an access condition for critical apps, not as an after-the-fact compliance check.

How SaaS discovery changes the governance model

SaaS discovery is a visibility function, but in identity terms it is also a control-enablement layer. If IT cannot see every application in use, then access reviews, deprovisioning, and license right-sizing are always incomplete. The article treats unmanaged SaaS as a governance blind spot because employee behaviour, seasonal workforce changes, and shadow adoption all create access paths outside central oversight. That is especially relevant when business units procure software faster than identity teams can classify it.

Practical implication: build application discovery into IAM and IGA workflows so hidden SaaS does not escape review and offboarding.

AI agent access and the limits of human-centric IAM

AI agent access behaves differently from human access because the actor can act at runtime inside a workflow rather than waiting for a person to sign in each time. That means access governance has to account for agentic workflows, credential reuse, and task-specific permissions that may not map neatly to user-centric IAM. The article signals that access control is no longer just about proving a person is who they claim to be. It is also about deciding what software actor is allowed to do, from which device, and under what trust conditions.

Practical implication: classify AI-driven access paths separately from human access and require explicit governance for agentic workflows.

NHI Mgmt Group analysis

Extended access management is a response to identity sprawl, not a replacement for IAM. The article describes a control problem where people, devices, SaaS, and AI-driven workflows all sit inside the same access perimeter. Traditional IAM and IGA were built around cleaner boundaries than most organisations now have. Practitioners should read this as a signal that access governance must span discovery, device trust, and workflow control together.

Human-centric access review models do not fully cover agentic access paths. The article’s own framing of AI agent access shows that software actors now sit inside ordinary access flows. That does not make them autonomous by default, but it does mean entitlement review, credential visibility, and device trust need to be explicit about non-human execution paths. IAM teams should separate human sign-in from machine and agent access in policy and reporting.

Trusted device status is becoming an identity control, not only an endpoint control. When access depends on whether the device is approved, the endpoint becomes part of the trust decision. That tightens the link between IAM, PAM, and endpoint governance. Organisations that still treat device posture as downstream telemetry will keep missing the point that access itself is conditioned by it.

SaaS discovery is now a prerequisite for lifecycle governance. If applications are not visible, then joiner, mover, and leaver processes cannot be complete. The article points to a governance gap in which access is granted faster than it is mapped, reviewed, or removed. Practitioners should treat unmanaged SaaS as an identity lifecycle problem, not just a procurement issue.

From our research:

• 1 in 4 organisations are already investing in dedicated NHI security capabilities, with an additional 60% planning to do so within the next twelve months, according to The State of Non-Human Identity Security.
• From our research: Only 5.7% of organisations have full visibility into their service accounts, according to the Ultimate Guide to NHIs.
• From our research: For lifecycle depth, review the NHI Lifecycle Management Guide for provisioning, rotation, and offboarding practices that close the governance gap.

What this signals

Extended access management will force IAM programmes to unify discovery, trust, and lifecycle work. Teams that still split SaaS visibility, device controls, and identity governance into separate workstreams will struggle to keep entitlement data current enough for review and offboarding. The operational signal is that access control is moving closer to runtime trust decisions, not just directory administration.

91.6% of secrets remain valid five days after notification in our research, which shows how slowly remediation can move once exposure is identified. That matters here because distributed access environments create more places where credentials, devices, and AI-driven workflows can persist longer than teams expect. Practitioners should watch for the same delay pattern in access removal and credential revocation.

The next governance boundary is not the app itself but the actor, the device, and the workflow together. As AI-driven operations spread, identity teams will need reporting that shows who or what is acting, from where, and under which trust condition. That is how lifecycle governance becomes operational rather than theoretical.

For practitioners

• Map unmanaged SaaS before tightening access policy. Feed application discovery into joiner, mover, and leaver workflows so hidden apps are included in access reviews, offboarding, and license rationalisation. If the app is not visible, the entitlement cannot be governed.
• Use device trust for high-risk access paths. Require device posture checks before granting access to sensitive systems, especially for remote users and shared environments. Treat trusted-device enforcement as a policy gate, not an audit log after sign-in.
• Separate human and agent access reporting. Track AI-driven workflows, service accounts, and human users in distinct access inventories so reviews do not collapse different actor types into one report. That makes it easier to spot non-human access that is over-broad or unmanaged.
• Tie credential sharing to explicit governance. Allow shared access only through approved processes with clear ownership, logging, and revocation steps. Shared credentials without lifecycle control create the same offboarding gaps the article warns about in fast-moving teams.

Key takeaways

• Extended Access Management reframes identity security around users, devices, SaaS, and AI-driven workflows at the same time.
• The main governance gap is not authentication alone, but incomplete visibility into unmanaged apps and non-human access paths.
• IAM teams should treat device trust, SaaS discovery, and agentic access reporting as one control plane, not separate projects.

Key terms

• Extended Access Management: An access governance model that brings users, devices, applications, and software actors into one control view. It is useful when identity no longer stops at login and must also account for device trust, SaaS sprawl, and machine-driven access paths.
• Device Trust: A policy approach that checks whether a device meets security requirements before access is granted. In practice, it makes endpoint posture part of the authorisation decision, which is important when remote work, shared devices, or unmanaged hardware can expose corporate systems.
• SaaS Discovery: The process of identifying applications in use across the organisation, including tools outside central IT control. It matters because access reviews, offboarding, and licensing decisions cannot be complete if security and identity teams cannot see the full application estate.
• Agentic Access: Access used by an AI-driven software actor that can execute actions inside a workflow rather than simply authenticate as a static integration. The governance challenge is to define what the actor may do, under which device or trust conditions, and how revocation works.

Deepen your knowledge

NHI governance, agentic AI identity, machine identity security, IAM, and secrets management are core topics in our NHI Foundation Level course, the industry's only accredited NHI security programme. If you are responsible for identity strategy or access governance, it is worth exploring.

This post draws on content published by 1Password: Extended Access Management for people, apps, devices, and AI. Read the original.