Shadow IT and shadow AI expose the access-trust gap

At a glance

What this is: NHI Mgmt Group's analysis, drawing on 1Password's blog post: Shadow IT and shadow AI expose a gap between who can use an app and whether that access is governed, visible, and revocable.

Why it matters: IAM, NHI, and PAM teams need to treat unsanctioned apps and AI tools as lifecycle and audit problems, not just user behaviour issues.

Context

Shadow IT and shadow AI describe employee use of unsanctioned applications and AI tools outside centrally managed access controls. In this article, the core issue is the access-trust gap that appears when identity, device, and application governance no longer move together.

That gap matters because legacy IAM and MDM assume the organisation can inventory, federate, and revoke access across the estate it manages. Once users bring their own apps, devices, and AI services into work, those assumptions stop holding and governance becomes fragmented.

Key questions

Q: What breaks when employees use apps and AI tools outside central IAM?

A: Central IAM breaks at the point where access was never enrolled in the first place. If an app, device, or AI service was adopted outside approved onboarding, identity teams cannot reliably recertify, revoke, or audit that access later. The failure is lifecycle reach, not just visibility.

Q: Why does shadow AI increase data exposure risk more than ordinary shadow IT in regulated environments?

A: Shadow AI is riskier because data is often entered through a prompt and ingested instantly, which can create immediate and sometimes irreversible exposure. Unlike a typical unsanctioned app, a consumer AI tool may retain inputs, lack a BAA, and bypass traditional network-based monitoring when users paste data locally or use desktop clients.

Q: What signals show that Shadow IT is becoming a governance problem?

A: Look for overlapping SaaS subscriptions, personal accounts used for work, OAuth grants to external apps, and licences left unused after 30 days. These are signs that ownership is unclear and lifecycle controls are failing. If you cannot explain why an app exists or who owns access, the problem is already governance-related.

Q: Should organizations prioritise SSO or lifecycle governance first?

A: Lifecycle governance should come first when access sprawl is already present, because SSO reduces login friction without correcting stale accounts or excessive access. Authentication convenience matters, but it does not remove entitlement drift. Organisations should treat SSO as an experience layer and lifecycle governance as the control layer.

Technical breakdown

Why legacy IAM breaks when apps are self-adopted

Legacy IAM works best when applications are pre-integrated, identities are federated, and access can be governed through a known control plane. Shadow IT breaks that model because employees create their own access paths outside central onboarding, so the organisation no longer has a complete authorization map. The result is not only poor visibility but weak lifecycle control, because offboarding and review processes cannot reliably reach systems that were never formally enrolled. This is why app discovery, access telemetry, and de-provisioning coverage matter more than broad policy statements.

Practical implication: inventory unsanctioned SaaS and web apps before you try to govern them.

How shadow AI changes the access problem

Shadow AI extends the same governance gap into model-based tools that can receive sensitive or proprietary data without corporate authentication or policy enforcement. The key issue is not simply model use, but the absence of controlled identity binding, approved data pathways, and auditable session context. When users interact through personal email, unmanaged accounts, or non-integrated authentication, security teams lose the ability to prove who accessed what, from where, and under which policy. That makes AI usage both a data protection issue and an identity governance issue.

Practical implication: require identity-aware access paths for AI tools that may process company data.

Why access-trust gap management is a lifecycle problem

The access-trust gap is ultimately a lifecycle problem because the organisation must govern discovery, approval, usage, and revocation across sanctioned and unsanctioned tools. If an app or AI service is invisible at onboarding, it is unlikely to be cleanly offboarded, recertified, or audited later. That creates persistence risk even when the original user leaves or changes role. In governance terms, the control failure is not a single missing policy; it is the absence of an enforceable lifecycle around every access path that reaches company data.

Practical implication: extend joiner-mover-leaver and recertification processes to unsanctioned applications and AI services.

Threat narrative

Attacker objective: The practical attacker objective is to exploit unmanaged access paths to retain data visibility, persistence, or unauthorized access beyond corporate control.

1. Entry occurs when employees adopt unsanctioned apps, personal devices, or shadow AI tools outside centrally managed onboarding.
2. Credential or identity context is then split across non-corporate email accounts, unmanaged authentication methods, or weakly governed app sign-ups.
3. Impact follows when sensitive data is processed, retained, or left accessible in systems that security teams cannot reliably monitor, revoke, or audit.

Breaches seen in the wild

• Okta support system breach 2023: A support service account credential saved in a personal Google profile let attackers take HAR files and hijack five Okta customers' sessions.

Read and download The State of NHI & AI Agent Breach Report 2026, covering 200+ breaches impacting Non-Human Identities including AI Agents.

NHI Mgmt Group analysis

Access-trust gap is now the governance problem, not a side effect: Shadow IT and shadow AI show that identity programmes fail when access can be created outside the sanctioned control plane. The issue is not just discovery, but whether the organisation can bind approval, authentication, and revocation to every place data moves. Practitioners should treat unmanaged access paths as first-class governance objects.

Legacy IAM and MDM assume managed enrolment, which modern work no longer guarantees: Those controls were built for environments where IT knew the app, device, and identity boundary in advance. That assumption breaks when users select their own tools and services at runtime, leaving central governance to catch up after the fact. The implication is that access management now has to start with visibility and lifecycle reach, not policy intent.

Shadow AI creates a data-governance blind spot that traditional access reviews miss: The problem is not simply that employees use AI, but that many of those sessions are not tied to corporate authentication or approved data pathways. That means recertification, logging, and offboarding can all fail to see the actual access relationship. Practitioners should regard unsanctioned AI use as an identity and data control failure, not just an acceptable-use issue.

Extended Access Management is best understood as a lifecycle response to unmanaged access paths: The category exists because organisations need discovery, access understanding, and de-provisioning across tools that never entered the formal IAM stack. That makes the real benchmark not SSO coverage alone, but whether the programme can govern access even when SSO is absent. Teams should judge their current controls by how well they handle the unauthorised edge, not the managed core.

What this signals

Access governance now has to follow user choice, not just system design: The practical boundary for IAM is no longer the set of apps IT provisioned. It is the broader set of tools employees adopt on their own, which means discovery and revocation need to cover the unmanaged edge as well as the sanctioned core.

Shadow AI turns identity gaps into data exposure paths: When AI tools are used without corporate authentication or approved integration, the organisation loses control over both access and data handling. That is why identity teams, not just security awareness teams, need to own the response.

Recertification only works when the estate is visible: Access reviews cannot remove risk from systems they cannot see. Teams that want meaningful governance should measure how much of their real application usage sits outside SSO, provisioning, and offboarding coverage.

For practitioners

• Map unsanctioned application sprawl Discover which SaaS, web, and AI tools employees are using outside approved onboarding so you can see where governance is missing.
• Extend offboarding to shadow access paths Make leaver processes reach non-federated apps and AI services that retain user data or sessions outside the managed identity stack.
• Require identity binding for AI use Block or limit AI tools that accept company data unless they are tied to corporate authentication and auditable access records.
• Review recertification coverage beyond SSO Test whether access reviews include apps and tools that were never integrated with SSO or central provisioning, because those are the gaps attackers and policy failures exploit.

Key takeaways

• Shadow IT and shadow AI expose a governance gap between employee access choices and the controls IAM teams can actually enforce.
• The main risk is persistence of access and data exposure in tools that were never enrolled in corporate lifecycle processes.
• Organisations need discovery, binding, and revocation across unmanaged apps and AI tools, not just stronger policy language.

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.
• Shadow IT: Shadow IT is the use of applications or services outside formal enterprise approval or visibility. In SaaS environments, it often includes department-purchased tools and unsanctioned integrations that create hidden identity, data, and access paths the security team cannot readily govern.
• Shadow AI: AI agents, copilots, or connected tools operating without full visibility or governance from security teams. Shadow AI becomes an identity problem when those systems authenticate with unmanaged tokens, service accounts, or OAuth apps that can reach production resources.
• 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.

Deepen your knowledge

NHI governance, identity lifecycle management, and secrets management 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.