SaaS and AI tools outside SSO expose unmanaged identity risk

At a glance

What this is: This is an analysis of the unmanaged stack problem, where shadow SaaS apps and AI tools sit outside SSO and create governance gaps that traditional IAM controls do not fully cover.

Why it matters: It matters because IAM, IGA, PAM, and NHI programmes increasingly have to govern access that is approved informally, forgotten operationally, and invisible to standard review cycles.

By the numbers:

• 52% of employees have downloaded apps without IT approval.
• The average estimated time to remediate a leaked secret is 27 days, despite 75% of organisations expressing strong confidence in their secrets management capabilities.
• When AWS credentials are exposed publicly, attackers attempt access within an average of 17 minutes.

👉 Read 1Password's webinar recap on governing SaaS apps and AI tools outside SSO

Context

The unmanaged stack is the set of SaaS applications and AI tools that sit outside standard identity controls, often because users adopt them faster than IT can govern them. In practice, that means shadow apps, ad hoc OAuth connections, and AI platforms with their own account models can bypass the visibility assumptions built into SSO-centric programmes.

For IAM teams, the problem is not just access sprawl but governance sprawl. When permissions are granted in one console, forgotten in another, and never revisited through normal lifecycle processes, organisations accumulate unmanaged non-human and human access paths that appear benign until audit, incident response, or cost review exposes them.

The article describes a typical enterprise tension rather than an edge case: people choose convenience first, and security teams inherit the resulting access graph later.

Key questions

Q: How should security teams govern SaaS and AI tools that sit outside SSO?

A: Start by inventorying every out-of-band access path, including OAuth grants, shared credentials, and vendor-specific admin accounts. Then assign ownership, review scope, and tie each access object to joiner-mover-leaver processes so it can be certified and retired like any other entitlement. Visibility without lifecycle control does not reduce risk.

Q: Why do unmanaged SaaS apps create identity risk even when users sign in legitimately?

A: Because valid authentication does not equal valid governance. A user can approve a third-party app once, then forget it while the grant remains active and continues to operate inside approved scopes. That turns convenience into standing privilege, which is hard for conventional monitoring to detect.

Q: What do teams get wrong about OAuth app governance?

A: They often treat OAuth as a one-time security decision instead of a lifecycle event. The better model is to review OAuth connections as living entitlements, verify whether elevated scopes are still needed, and revoke grants that no longer map to an active business purpose.

Q: Who is accountable when unsanctioned AI tools or shadow SaaS access cause an incident?

A: Accountability usually sits with the business owner who allowed the tool, the identity team that failed to govern the grant, and the security function that did not maintain visibility. The clearest answer comes from documenting ownership at the moment access is created and keeping revocation authority explicit.

Background and context

Why SaaS outside SSO becomes an identity blind spot

When an application cannot be mediated through SSO, identity controls lose the normal hooks for authentication policy, session oversight, and lifecycle enforcement. The result is not simply “shadow IT”, but a parallel entitlement layer where access may exist in personal vaults, browser extensions, or vendor-specific admin consoles. That matters because discovery, review, and revocation each depend on different control planes. If teams cannot continuously enumerate those access paths, they cannot reliably certify them, scope them, or retire them on schedule.

Practical implication: build discovery and review into the same governance loop so unmanaged apps are not left outside certification and offboarding.

OAuth connections and standing privilege in unmanaged apps

OAuth creates valid, scoped access without requiring password reuse, which is useful operationally but risky when permissions are granted once and never revisited. In an unmanaged stack, the threat is not failed login attempts but authenticated requests that look legitimate to conventional detection tools. Standing privilege appears when tokens and grants outlive the business need that justified them. Once a third-party app is compromised, the attacker does not need to break authentication. They inherit the approved scope and act inside it.

Practical implication: treat OAuth grants as governed entitlements and review them with the same rigor as any other privileged access path.

AI tool governance needs lifecycle control, not just approval

Approved AI platforms can still sit outside central identity governance if onboarding, role assignment, usage review, and offboarding happen in separate tools or informal workflows. That creates a lifecycle gap even when the tool itself is sanctioned. The article points to a broader pattern: AI access is often managed transactionally, not operationally. For identity teams, the key issue is whether AI accounts, roles, and usage data are tied to a lifecycle process that can prove who owns the account, who can revoke it, and when access should end.

Practical implication: align AI tool governance with joiner-mover-leaver controls so account ownership and removal are not left to manual cleanup.

NHI Mgmt Group analysis

Unmanaged access is now a lifecycle problem, not just a discovery problem. The article shows that the real failure is not whether a tool exists outside SSO, but whether it ever enters the governance lifecycle at all. Once access is approved informally and then forgotten, traditional review processes never see it again. That means IAM, IGA, and NHI controls have to govern the same access path from first grant to final revocation, or they are operating on an incomplete estate.

Standing privilege exposure is the specific failure mode this article exposes. OAuth and similar access paths are often granted for convenience, then left in place after the original task ends. That is not a generic shadow IT issue. It is a clear example of entitlements outliving accountability, which creates a standing privilege window across SaaS and AI tools. Practitioners should treat that as a governance defect, not an inventory nuisance.

Unmanaged stack sprawl is becoming a cross-domain identity problem. Human users drive adoption, but the access objects themselves behave like non-human identities once they are created and left to persist across tools. That means the same governance discipline must cover human choice, NHI-style credential persistence, and AI tool lifecycle control in one model. Security teams that separate those concerns will continue to miss the connections between convenience, persistence, and abuse.

Policy without continuous visibility is not governance. The article’s central lesson is that teams cannot certify what they cannot see and cannot revoke what they cannot enumerate. That is why unmanaged SaaS and AI access belongs in identity governance, not only in security operations. Practitioners should reframe the issue as control-plane completeness, because incomplete control planes produce incomplete accountability.

Named concept: unmanaged stack drift. This is the cumulative state in which shadow SaaS, OAuth grants, personal vaults, and AI tool accounts accumulate faster than governance can absorb them. The drift is not just technical sprawl. It is the gradual separation of business use, account ownership, and revocation authority, which leaves identity programmes managing yesterday’s access while users keep creating today’s.

From our research:

• The average estimated time to remediate a leaked secret is 27 days, despite 75% of organisations expressing strong confidence in their secrets management capabilities, according to The State of Secrets in AppSec.
• Only 44% of developers are reported to follow security best practices for secrets management, exposing a significant developer behaviour gap.
• That governance gap is one reason teams should also study Ultimate Guide to NHIs , Lifecycle Processes for Managing NHIs for lifecycle control patterns that apply across identity types.

What this signals

Unmanaged stack drift is becoming a predictable programme failure mode. The more apps and AI tools users adopt outside SSO, the more identity teams need a control model that spans discovery, entitlement review, and offboarding in one workflow. That shift is already visible in adjacent secrets research, where delayed remediation shows how quickly confidence can outrun actual control. For practitioners, the signal is clear: visibility without retirement authority will not hold.

As AI tool adoption expands, identity programmes will need to treat browser-based sign-ins, OAuth grants, and shared vault access as first-class governance objects. The programme implication is not to block every tool, but to decide which access paths can be tolerated, which require tighter review, and which should be retired because they no longer fit the operating model. The teams that codify that distinction now will have cleaner audit evidence later.

For practitioners

• Inventory unmanaged SaaS and AI access paths Pull together OAuth grants, browser-extension sign-ins, shared vault credentials, and vendor-specific admin accounts into one reviewable list so you can see where access lives outside SSO and who owns each grant.
• Classify access by lifecycle state Tag each discovered app or tool as approved, tolerated, or orphaned, then assign a named owner and retirement date so the access object can move through joiner-mover-leaver controls instead of remaining in limbo.
• Review OAuth scopes and token lifetime Compare each third-party connection against actual business need, remove elevated scopes where possible, and shorten token lifetime where the tool does not require persistent access.
• Bring AI tools into access governance Require onboarding, role assignment, usage review, and offboarding for AI platforms in the same workflow used for other identity objects, including documented approval for shared or high-risk accounts.

Key takeaways

• The unmanaged stack is an identity governance problem because access can be approved, forgotten, and still remain active outside central control.
• The clearest failure mode is standing privilege persistence, where OAuth grants and shared logins outlive the business need that created them.
• Practitioners should govern shadow SaaS and AI tools as lifecycle objects, with ownership, review, and revocation built into the same process.

Key terms

• Unmanaged Stack: The unmanaged stack is the collection of SaaS apps, OAuth grants, shared credentials, and AI tools that operate outside central identity controls. It creates governance gaps because access may be real, active, and risky even when it never appears in normal SSO review or lifecycle workflows.
• Standing Privilege: Standing privilege is access that remains available after the business reason for it has ended. In unmanaged SaaS and AI environments, it often appears as forgotten OAuth scopes, stale shared logins, or persistent admin rights that were granted for convenience and never retired.
• OAuth Grant Lifecycle: OAuth grant lifecycle is the process of approving, reviewing, scoping, and revoking third-party application access over time. It matters because a valid grant can become a long-lived entitlement if teams do not track ownership, scope drift, and offboarding with the same discipline used for other identities.
• Shadow IT: Shadow IT is technology adopted by employees without formal IT approval or governance. In identity terms, the risk is not just unauthorized software use, but access paths that accumulate outside review, making them difficult to certify, monitor, and remove when the business need disappears.

Deepen your knowledge

Unmanaged SaaS governance and AI tool lifecycle control are core topics in our NHI Foundation Level course, the industry's only accredited NHI security programme. If you are building a programme to close the SSO gap, it is worth exploring.

This post draws on content published by 1Password: The unmanaged stack, governing SaaS apps and AI tools outside SSO. Read the original.