1Password’s cloud 100 recognition spotlights the access-trust gap

At a glance

What this is: This is 1Password’s Cloud 100 recognition note, and its key finding is that unmanaged apps, unmanaged devices and AI agents are widening the access-trust gap beyond legacy IAM coverage.

Why it matters: It matters because IAM, IGA and PAM teams now have to govern access paths that sit outside classic SSO and device-management assumptions, including human and AI-driven access to SaaS.

Context

Extended access management is 1Password’s label for the gap between where work now happens and where older identity controls still operate. The article argues that people and AI agents are using more tools, more devices and more unsanctioned applications than traditional security programmes were designed to govern.

The core governance problem is not recognition on a cloud ranking. It is that access decisions are being made across unmanaged surfaces, while trust signals remain fragmented across identity, application and device layers. That makes this a cross-domain identity issue, not just a SaaS access problem.

Key questions

Technical breakdown

Why legacy IAM leaves an access-trust gap

Legacy IAM was built around a narrower enterprise boundary: federated sign-in, managed devices and sanctioned applications. When work moves into unmanaged apps, BYOD-like devices and AI-mediated access, the control plane no longer sees the full path from identity to resource. The result is not just weaker authentication. It is incomplete governance over where trust is established, how it is consumed and where it silently disappears. That is why access management now has to account for app reach, device health and identity type together rather than treating them as separate hygiene domains.

Practical implication: map where SSO stops and where unmanaged access begins, then close those blind spots before expanding AI or SaaS adoption.

How AI agents change identity and access assumptions

The article’s most material signal is that AI agents are not just extra users. They alter the access model because they can operate with more autonomy, move across more tools and interact with systems at machine speed. That changes the governance problem from one of user convenience to one of runtime trust. If the actor can be an agent, a human or both, then identity assurance, authorization scope and device trust all need to be evaluated at the point of access, not inferred from a directory record alone.

Practical implication: define how AI-agent access is authenticated, bounded and reviewed before agents are allowed to interact with sensitive workflows.

Extended access management sits between IAM, IGA and MDM

Extended access management is best understood as a control layer that tries to cover the surfaces legacy IAM, IGA and MDM do not fully reach. IAM proves who signs in, IGA governs entitlement decisions over time, and MDM manages device posture for enrolled endpoints. But none of those alone guarantees governance across unsanctioned apps, transient device contexts and non-human actors. The architectural point is that modern access trust depends on stitching these controls together rather than assuming any single domain owns the whole problem.

Practical implication: assess whether your current identity stack can actually govern unsanctioned apps, unmanaged devices and non-human access in one policy model.

NHI Mgmt Group analysis

The access-trust gap is now a governance boundary problem, not a tooling gap. The article shows that work no longer stays inside the boundaries assumed by legacy IAM, IGA and MDM. When access happens through unmanaged devices and unsanctioned apps, governance fails because the programme does not see the full trust chain. Practitioners should treat this as a boundary redesign issue, not a feature request.

AI agents turn access scope into a dynamic governance problem. AI agents are not simply another workload or another user group. They can shift between apps, devices and tasks with more autonomy than traditional identity models assume, which makes static entitlement thinking weaker. The implication is that access policy must be evaluated against actor behaviour, not just identity labels.

Extended access management names a real control pattern, but the problem space is broader than one category label. The useful insight is not the marketing term itself, but the fact that trust now has to follow the session across identities, devices and applications that classic stack boundaries do not cover. That pushes IAM teams, IGA teams and endpoint teams toward a shared governance model instead of isolated ownership.

Unmanaged access is the new identity perimeter. The article’s central signal is that security teams should stop treating SaaS sprawl, unmanaged devices and AI-mediated access as separate exceptions. They are now the normal operating environment. Practitioners who do not redesign controls around that perimeter will keep governing the wrong boundary.

From our research library:

• 19% of organisations give AI systems dramatically more access than human employees, nearly one in five granting unrestricted privilege, according to the 2026 Infrastructure Identity Survey.
• 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.
• Read next: Zero Trust for AI Agents

What this signals

Unmanaged apps and unmanaged devices are now first-class identity problems. The practical shift is that identity governance has to extend beyond directory-backed sign-in into the places where work actually happens. If policy only covers sanctioned systems, the programme is already blind to part of the access estate.

AI agents are forcing access reviews to become context-aware. Reviews that only ask whether an identity exists no longer answer the more important question of whether the actor should be allowed to operate in that environment at all. 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.

Access-trust gap: the control model breaks where identity, device and application governance do not meet. That gap should shape roadmap priorities for IAM, IGA and endpoint teams together, not in sequence. The strongest programmes will define which access paths need shared ownership before AI adoption expands further.

For practitioners

• Define the access-trust boundary Map where your current IAM, IGA and MDM controls stop seeing access, especially in unmanaged apps and devices.
• Inventory AI-mediated access paths Identify which business workflows now involve AI agents, then classify the access they use, inherit or proxy.
• Tighten governance for unsanctioned apps Require explicit approval, logging and periodic review for application access that sits outside the sanctioned application estate.
• Separate human and non-human trust decisions Do not let directory membership alone authorise either human or agent access when device posture and app risk are unknown.

Key takeaways

• Legacy identity controls no longer cover the full access path when work moves across unmanaged apps, devices and AI-mediated interactions.
• The biggest risk is governance blind spots at the boundary between IAM, IGA and endpoint control, not just weak sign-in.
• Practitioners should map where access trust is established and revoke assumptions that directory membership alone is enough.

Key terms

• Access-trust gap: The gap between having a policy and actually enforcing it when access is requested. It appears when compliance, HR, or training data sits in separate systems from the access decision, allowing users to reach resources even though a required condition has not been met.
• Extended access management: Extended access management is a control model that covers access beyond traditional employee login, including SaaS, service identities, and AI agents. It tries to unify trust, entitlement, and device context so organisations can govern access wherever work actually happens.
• Unmanaged Access: Any access to company data that does not pass through the organisation's primary control and logging stack. In practice, this includes devices, apps, or identities that can authenticate or reach data without being fully visible to IAM, MDM, or audit workflows.
• AI-mediated access: Access that is initiated or transformed through an AI tool rather than a direct human action alone. In practice, the identity controls underneath the tool still determine who can reach data, what the tool can expose, and whether the resulting activity is auditable.

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.