Identity security, AI governance and third-party risk are converging

At a glance

What this is: This SailPoint conversation with KPMG argues that identity governance is converging with AI governance and third-party risk as organisations confront cloud sprawl, autonomous access, and hybrid operating models.

Why it matters: IAM and IGA teams now have to govern people, vendors and AI-driven actors under the same access-risk lens, or the control model will lag the way organisations actually operate.

By the numbers:

• 60% of breaches being identity-based

Context

Identity security is no longer only about granting and reviewing access. In this article, SailPoint frames it as a governance problem tied to business continuity, compliance and digital transformation, with the pressure increasing as cloud adoption and platform integration expand the access surface.

The central issue for identity teams is that third-party access and AI-enabled access now sit inside the same operating environment as human users. That means governance has to cover not just who can log in, but which external parties and autonomous agents can reach sensitive data, when, and under what policy boundary.

Key questions

Q: What breaks when third-party access is not lifecycle managed?

A: Access outlives accountability. When vendor credentials are not tied to a clear offboarding process, old support paths, dormant accounts, and overbroad entitlements remain available after the business need has ended. That creates a standing exposure window that attackers can exploit and auditors will struggle to explain.

Q: Why do autonomous agents change identity governance more than chatbots do?

A: Because the risk moves from generated content to real access. A chatbot can leak information through prompts or output, but an autonomous agent can also reach tools, systems, and data stores with credentials. That turns identity scope, ownership, and revocation into the primary controls, not model output filters.

Q: Should organisations prioritise cloud identity governance before expanding privileged access controls across applications?

A: Yes. Organisations should start with a governance layer that gives them visibility into identities, entitlements, and application access, then extend privileged controls where the highest risk exists. That sequence helps teams establish policy, approvals, and review discipline before they manage elevated access at scale. It also makes later PAM expansion easier to operationalise.

Q: What is the difference between runtime authorization and access reviews for agents?

A: Access reviews look backward at entitlements already granted, while runtime authorization controls the action before it happens. For AI agents, that distinction matters because the risky decision often occurs after access is approved, when the agent selects a tool or data source that changes the scope of the session.

Technical breakdown

Why third-party identity exposure changes the governance model

Third-party access changes identity governance because the trust boundary is no longer limited to employees and contractors. When vendors connect into hybrid environments, their credentials, tokens and delegated access often outlive the business context that created them. That creates a governance problem, not just an access problem, because accountability depends on lifecycle control, owner visibility and timely offboarding. In practice, teams need a clear inventory of external identities, the systems they touch and the business justification for each relationship. Without that, access reviews become a paper exercise rather than an actual control.

Practical implication: treat third-party accounts as governed identities with explicit ownership, review, and removal paths.

AI governance and autonomous agents in identity programmes

AI governance enters identity security when organisations give autonomous agents access to data or administrative workflows. The important distinction is that these actors are not just automated scripts. If they can initiate actions and consume privileges dynamically, then the identity programme has to govern runtime access scope, not only provisioning. That shifts attention toward time-bound access, policy enforcement and auditability at execution time. It also changes how teams think about overprovisioning, because an agent that can act faster than a review cycle can create exposure before a human ever sees it.

Practical implication: govern AI agents as active identities with runtime access limits and auditable decision points.

Identity security, business continuity and compliance are converging

This article reflects a broader trend: identity governance is becoming a control plane for resilience, not a back-office access function. Business continuity depends on knowing which identities can fail, which third parties can interrupt critical services, and which AI-driven actors could expand risk during change. Compliance follows the same logic, because evidence of correct access must stand up across complex, hybrid environments. The practical effect is that IAM, IGA and PAM teams need to align around access assurance rather than separate tool ownership.

Practical implication: align access governance, resilience and audit evidence around one identity operating model.

Breaches seen in the wild

• Salesloft OAuth token breach: hackers stole OAuth tokens to access Salesforce data via Salesloft.
• BeyondTrust breach 2024: A stolen BeyondTrust Remote Support API key let a China state-sponsored actor reset accounts and reach US Treasury workstations in 2024.

Read our 52 NHI Breaches Analysis report for a comprehensive view of breaches impacting Non-Human Identities including AI Agents.

NHI Mgmt Group analysis

Identity governance is becoming the control plane for enterprise resilience. The article shows that identity has moved from a permissions topic to a business continuity topic because cloud integration and external access now shape operational continuity. That is a material shift for governance models built around periodic access review alone. The practitioner conclusion is that identity programmes must be evaluated by how well they preserve service continuity under change, not just by how many reviews they complete.

Third-party access is no longer a perimeter exception, it is part of the identity estate. SailPoint's framing reflects a reality many programmes still under-handle: vendor access is often operationally embedded but governance-light. Once external parties can touch sensitive systems, the question is lifecycle control, ownership and offboarding discipline, not simply authentication strength. The practitioner conclusion is that external identities need the same accountability model as internal ones.

Autonomous agents create an access governance problem that human-centric review models were not built to absorb. Access review was designed for privileges that persist long enough to be observed and certified. That assumption fails when autonomous agents can receive, use and discard access within a short runtime window, often before a review cycle can register the state. The implication is that governance has to move from periodic attestation to runtime control and issuance-time policy.

AI governance and identity security are converging because both are now deciding who or what can act inside the enterprise. The article's duality of AI is important: organisations want AI to improve security operations while also constraining AI-driven access. That creates a governance overlap between AI risk, identity lifecycle and privilege management. The practitioner conclusion is that identity leaders should treat AI access as a governance domain, not an isolated technology project.

Identity security needs a named concept: access-rights provenance. The article points to a growing need to prove not just that access exists, but why it exists, who approved it and whether that approval still matches the current relationship. In hybrid environments, especially with third parties and AI-driven actors, provenance becomes the only durable evidence of legitimate access. The practitioner conclusion is that identity programmes should prioritise evidence of entitlement origin and current justification.

From our research library:

• Only 44% of organisations have implemented any policies to manage their AI agents, despite 92% agreeing that governing AI agents is critical to enterprise security, according to the 2026 Infrastructure Identity Survey.
• Read next: Ultimate Guide to NHIs — Key Research and Survey Results

What this signals

Access-rights provenance: identity programmes now have to prove why access exists, not just that it was approved once. In environments where cloud integrations, third parties and AI-driven actors all operate inside the same governance boundary, provenance becomes the practical evidence layer for audit and operational trust.

The strongest programmes will stop treating AI access as a side issue and start governing it as a first-class identity domain. Only 44% of organisations have implemented any policies to manage their AI agents, despite 92% agreeing that governing AI agents is critical to enterprise security, according to the 2026 Infrastructure Identity Survey. That gap is where unmanaged risk accumulates.

For practitioners

• Audit third-party access lifecycles Inventory vendor, partner and outsourced accounts across connected platforms, then tie each one to an owner, business purpose and offboarding trigger.
• Separate AI agent access from human access Create a distinct governance path for autonomous agents so their entitlements, approvals and logging are not forced into human access review cadences.
• Rework access reviews for hybrid environments Focus recertification on active business relationships, system touchpoints and stale delegated access, not on identity lists alone.
• Use runtime controls for sensitive agent activity Limit what AI-driven actors can reach at execution time, and require audit evidence that the access decision matched policy at issuance.
• Align identity evidence with compliance reporting Capture who approved access, why it was granted and when it should expire so audit evidence survives cloud and third-party complexity.

Key takeaways

• Identity governance is being pulled into continuity, compliance and transformation decisions as external access and AI-driven actors expand the control surface.
• The article cites 60% of breaches as identity-based, which reinforces that access governance is now a primary risk-management function, not an administrative one.
• Practitioners should tighten lifecycle ownership, runtime access limits and audit evidence so hybrid identity programmes can prove legitimacy under change.

Key terms

• Access-rights provenance: The evidence trail that explains why an identity has a permission, who approved it and whether the justification still holds. In hybrid and AI-enabled environments, provenance is the difference between a controlled entitlement and an unexplained standing risk.
• Third-Party Identity: An identity issued to a partner, vendor, contractor, or external service that can access internal systems. These identities often sit outside normal employee governance and can become persistent trust paths if they are not reviewed, expired, and revoked on schedule.
• Runtime Access Control: Policy enforcement that evaluates an identity's action at the moment it tries to do something, rather than only at login or provisioning time. For AI agents, this is critical because they can chain actions dynamically and exceed their intended scope without a new authentication event.
• Identity Governance: Identity governance is the set of controls that defines who approves access, who owns it, how it is reviewed, and when it is removed. In practice, it turns identity management from a deployment task into a durable control system that can withstand audits, organisational change, and operational growth.

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.