Independent analysis by NHI Mgmt GroupSource article: Oak, “Identity Governance in Financial Services, for Every Identity Type” (November 15, 2026)

TL;DR: Oak shows that financial services identity governance must cover human, machine and AI agent identities under PCI DSS, DORA, SOX and NYDFS, because review cadence, segregation of duties and examiner evidence all depend on a complete access record. Separate inventories leave regulated access invisible, so governance has to move to one live model.


At a glance

What this is: NHI Mgmt Group's analysis, drawing on Oak's blog post: Financial identity governance now has to certify human, machine and AI agent access on one record that supports review cadence, segregation of duties and audit evidence.

Why it matters: Banks, insurers and payments firms cannot rely on separate inventories when regulated access, third-party reach and machine identities all create audit and control exposure.

👉 Read Oak's analysis of identity governance for financial services


Context

Identity governance in financial services fails when access is split across people, machines and AI agents that live in different systems of record. Regulators and auditors care about effective access, review cadence and evidence, not where a team happened to store the identity.

Financial firms need to govern the full access population, including contractors, service accounts and agent identities, because the same control obligations apply even when the account is not tied to a person. That is why financial identity governance has become a lifecycle and evidence problem as much as an access problem.


Key questions

Q: How should teams certify effective access in financial systems?

A: Teams certify effective access by reviewing the access a user can actually exercise through direct entitlements, roles, groups and inheritance, not just the permissions listed on paper. The review should confirm whether access is still needed, whether privileged access is justified, and whether dormant or inherited rights should be removed or re-approved.

Q: What is the best way to govern service accounts in a bank?

A: The best way to govern service accounts is to assign each one an internal owner, a defined purpose, and a documented system scope. Then control credentials with rotation, limit interactive logon, remove unused accounts, and review entitlements regularly so each account stays tied to a specific workload or application.

Q: How do payment workflows create segregation of duties risk?

A: Payment workflows create segregation of duties risk when one person can both initiate and approve a transaction, because that combines conflicting controls in a single path. The risk is usually managed with separate roles, approval thresholds, dual authorisation, and audit trails that show who created, reviewed and released each payment.

Q: Why do third-party accounts create audit problems?

A: Third-party accounts create audit problems because they often lack a named owner, a full inventory, and a clear business justification. Without those controls, it is hard to prove who is accountable, what access the account has, when it was last reviewed, and whether it should still exist.


Technical breakdown

Effective access across nested groups and inherited rights

Effective access is the permissions a subject can actually use after groups, inherited roles and impersonation paths are resolved. In financial environments, that matters because a clean entitlement list can still hide a toxic combination once the system calculates the real access path. Segregation of duties failures usually appear only after those layers are flattened. Auditor scrutiny therefore belongs on resolved access, not on the surface inventory alone. See the IAM and IGA Basics.

Practical implication: Review the resolved access path for every privileged or payment-related identity before certifying it.

Continuous reviews for human, machine and AI agent identities

Continuous review means access decisions are triggered by change, not only by a scheduled campaign. That matters when service accounts, contractors and AI agent identities accumulate permissions outside the cadence of a human recertification cycle. A live model can use last use, sensitivity and ownership to prioritise review, while the old campaign model often leaves machine identities outside the process until an examiner asks for proof. See the NHI Lifecycle Management Guide.

Practical implication: Move review triggers to identity change events so machine and agent access is assessed when risk appears.

Segregation of duties in payment workflows

Segregation of duties breaks when one identity can both create and approve a payment, even if those rights sit in separate systems. Financial firms often miss this because entitlement lists do not show the combined effect of role inheritance, nested groups and cross-system privileges. The control failure is not the individual right, but the merged authority that appears only after access is correlated. See the Segregation of Duties (SoD) Guide.

Practical implication: Correlate create and approve rights across systems before certifying any finance role.


Threat narrative

Attacker objective: The objective is to preserve or exploit access that should have been limited, reviewed or removed before it creates regulatory, financial or control failure.

  1. entry: A user, contractor or service account gains access that appears acceptable in one inventory but becomes excessive once other systems and inherited rights are resolved.
  2. escalation: The identity accumulates effective permissions across payment, operational or administrative workflows without a single control noticing the combined authority.
  3. impact: The organisation certifies risky access, weakens segregation of duties and creates audit evidence that does not match the real access state.
  • Dropbox Sign breach 2024: A compromised back-end service account gave attackers Dropbox Sign customer data, including API keys, OAuth tokens and MFA information.
  • Kubeflow cryptomining attacks 2020: Exposed Kubeflow dashboards and a notebook's mounted Kubernetes service account let attackers run cryptominers on tens of ML clusters.

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

Financial identity governance breaks when access is split across records that do not resolve to one accountable view. The core failure is not lack of review intent, but lack of a single model that can answer who and what has access once inheritance, delegation and third-party paths are flattened. Financial firms need one governance truth set, or they will keep certifying the wrong thing.

Regulatory cadence is a floor, not a programme design. PCI DSS, NYDFS and DORA each define access expectations, but the practical question is how a firm keeps evidence current between those formal checkpoints. A live access model matters because the examiner does not inspect policy intention, only whether the reviewed population matches the controlled population.

Machine identities now force IGA and NHI governance into the same operating model. When service accounts outnumber people, the traditional division between IGA for humans and spreadsheets for machines stops working. The Ultimate Guide to NHIs becomes relevant here because the financial-services problem is no longer only entitlement review, but ownership, lifecycle and evidence across every identity type.

SoD failures in finance are usually correlation failures, not entitlement failures. A payment approver can look clean in isolation and still be toxic once the initiating right is resolved from another role or system. That means the control problem is effective access, not just explicit access, and practitioners need reviewer workflows that understand combined authority.

Third-party access becomes an internal governance problem as soon as the vendor account reaches regulated systems. The control obligation does not stop at procurement or contract terms. Once a contractor, processor or integration account can reach core financial data, it needs the same ownership, review and offboarding discipline as any internal identity.

From our research library:

What this signals

One live identity graph is now the practical baseline for financial governance. Separate inventories cannot support effective review when the same account can touch payments, customer data and administration from different systems. The programme implication is simple: if the record does not resolve to effective access, it does not support certification.

Audit evidence is shifting from assembled packets to continuously maintained proof. Financial firms need reviewer context, revocation status and management acknowledgment in the same control record, because examiners increasingly want to see how decisions were made, not just that a review happened. That turns evidence into an operational output of governance, not a year-end exercise.

Ownership is the control that ties machine and agent access back to governance. A service account or AI agent without a human owner is operationally convenient and administratively fragile. When the owner is missing, offboarding, review and exception handling all degrade at the same time.


For practitioners

  • Map effective access, not just assigned access Resolve nested groups, inherited rights and impersonation paths before certification so reviewers see the authority an identity can actually use.
  • Move to event-driven access reviews Trigger review when an identity changes role, gains privilege, goes dormant or is onboarded as a third party, instead of waiting for the next campaign.
  • Treat contractor and vendor accounts as owned identities Assign an internal owner to every external account and require the same review and revocation discipline used for employee access.
  • Correlate payment rights before approval Check create, approve and override rights together across finance systems so one person cannot hold both sides of the payment control.
  • Build examiner-ready evidence from the live graph Preserve reviewer context, last use, revocation proof and management acknowledgment in the same record that drives certification decisions.

Key takeaways

  • Financial identity governance fails when people, machines and AI agents are reviewed in separate systems that cannot show effective access.
  • Regulated access in banks, insurers and payments firms needs one live record that supports review cadence, segregation of duties and audit evidence.
  • Ownership, lifecycle and evidence have become the decisive controls for external accounts and machine identities in financial services.

Related NHI Mgmt Group resources: NHI Lifecycle Management Guide · Segregation of Duties (SoD) Guide · Service Account Security Guide · Ultimate Guide to NHIs: Regulatory and Audit Perspectives


Standards & Framework Alignment

This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.

OWASP Non-Human Identity Top 10 addresses the attack and risk surface, while NIST CSF 2.0, CIS Controls v8 and NIST SP 800-53 Rev 5 set the governance and control requirements practitioners need to meet.

FrameworkControl / ReferenceRelevance
OWASP Non-Human Identity Top 10NHI-05 — Overprivileged NHIMachine identities in financial services often accumulate access beyond their role and owner controls.
NHI-01 — Improper OffboardingExternal and service identities can outlive the business need when offboarding does not remove access.
Recommendation — Review machine identity scope against NHI-05 and remove excess permissions before the next certification cycle. Tie offboarding to NHI-01 so every departed role, vendor or integration loses access immediately.
NIST CSF 2.0PR.AA-05 — Access Permissions, Entitlements and AuthorizationsThe post centres on who can access regulated systems and whether those entitlements stay appropriate.
Recommendation — Apply PR.AA-05 to keep entitlements limited, reviewed and aligned to job function across all identity types.
CIS Controls v8CIS-5 — Account ManagementFinancial identity governance depends on account ownership, lifecycle control and timely revocation.
Recommendation — Use CIS-5 to track account ownership and remove stale access from human and machine identities.
NIST SP 800-53 Rev 5IA-5 — Authenticator ManagementService accounts and regulated users both depend on managed authenticators and revocation discipline.
Recommendation — Apply IA-5 to control credential lifecycle for human and non-human identities that access sensitive financial systems.

Key terms

  • Effective Access: The actual permissions an identity can exercise after inheritance, nested groups, delegation, and object-level controls are evaluated. In Active Directory, effective access is more useful than direct membership because it reveals the true operational reach of a service account.
  • Segregation of Duties: Segregation of Duties is a control principle that prevents one person or role from combining incompatible permissions that could create fraud, error, or undetected change. In ERP environments, it must account for roles, transactions, approvals, and compensating controls across business processes.
  • Machine Identity: The digital identity of a machine, device, or workload, such as a server, container, or VM, used to authenticate it within a network. Sometimes used interchangeably with NHI, though NHI is the broader category.
  • Joiner Mover Leaver: Joiner Mover Leaver is the identity lifecycle process for creating, changing, and removing access as people enter, change roles, or leave an organization. It governs provisioning, modification, and deprovisioning across systems, ensuring access matches current job needs and reducing orphaned accounts, privilege creep, and residual access risk.

What's in the full article

Oak's full blog post covers the operational detail this post intentionally leaves for the source:

  • Implementation detail for connecting the live graph to existing IGA and ticketing workflows
  • Operational guidance for reviewer context, revocation proof and management acknowledgment
  • Examples of how Oak maps human, machine and AI agent identities into one model
  • Details of integrations and connectors for systems the IGA cannot currently reach

👉 Oak's full post covers the access review cadence, SoD examples and evidence model in more detail.

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.
NHIMG Editorial Note
Published by the NHIMG editorial team on November 16, 2026.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org