Join our Newsletter — 33% off our NHI Course
Home› FAQ› Governance, Ownership & Risk› How should security teams govern third-party app access…
Governance, Ownership & Risk

How should security teams govern third-party app access in Salesforce environments?

← Back to all FAQ
By NHI Mgmt Group Editorial Team Updated October 30, 2026 Domain: Governance, Ownership & Risk

Security teams should inventory every connected app and service, vet non-marketplace integrations before approval, and continuously review the permissions each connection receives. The goal is to remove unused connections, enforce least privilege, and monitor for unusual behavior such as unexpected geolocation, IP, or user-agent changes that can signal compromised access or misuse.

How Salesforce third-party app access should be governed

Governance starts with a complete inventory of every connected app, integration user, and external service that can reach Salesforce data. That inventory should distinguish approved marketplace apps from direct OAuth or API integrations, because the approval path, ownership, and review depth are often different. A practical baseline is to treat every new connection as a privileged access decision, not a routine IT request.

The strongest control is to make approval conditional on business purpose, data scope, and permission scope. If an app only needs read access to a limited object set, it should not receive broad API or administrative entitlements. That same principle should apply to dormant integrations, where unused access should be removed rather than kept “just in case.”

For a good operational model, use a formal review process for Salesforce connected apps and integration users, including permission review, token lifecycle checks, and revocation when the business no longer needs the connection. NHIMG’s How to review Salesforce connected apps and integration users is a direct fit for that workflow.

How to separate legitimate integrations from high-risk app sprawl

Not every third-party app deserves the same level of trust. Non-marketplace integrations, especially those using broad OAuth consent or long-lived tokens, need closer scrutiny because they can bypass normal user-by-user controls once approved. The key governance question is whether the app is scoped tightly enough to justify its existence, and whether an owner can explain why it needs the access it has.

Inventory work should also look for shadow SaaS and unsanctioned applications that were connected outside the normal procurement or security review path. Those apps often survive because they still “work,” not because they are still needed. Linking discovery to consent review helps teams find integrations that were approved once and then forgotten.

Where discovery is a recurring problem, use a dedicated app discovery process that combines identity logs, OAuth grant review, and other telemetry sources. NHIMG’s Shadow SaaS and Unsanctioned App Discovery Guide supports that broader inventory and triage approach.

In practice, the “is this app allowed?” question should be answered before the “does it technically work?” question. If an integration cannot be tied to a business owner, a data purpose, and a current support path, it should be treated as a candidate for removal.

What Salesforce teams should watch after approval

Approval is only the starting point. Continued governance depends on watching for changes that suggest the app, token, or user session is no longer behaving as expected. A sudden change in source IP, geolocation, or user agent is especially important when a connection normally originates from a fixed service environment, because that can indicate token theft, replay, or misuse.

This is also where permission drift matters. An app that started with a narrow use case may later receive broader object, API, or admin access because a business team asked for a quick exception. Those exceptions accumulate quietly unless someone is explicitly reviewing the effective permissions that each connection can still exercise.

Teams should also assume that compromise of a third-party integration can become a Salesforce data issue very quickly if the token can read records, export reports, or call downstream APIs. Past breach patterns show that stolen OAuth material can be enough to reach CRM data without needing a password at all. NHIMG’s Salesloft OAuth token breach is a useful example of that failure mode, and Gainsight Salesforce breach 2025 shows why long-lived tokens deserve special scrutiny.

Risk and Threat Considerations

Third-party app access becomes risky when the organisation trusts the app more than it understands the app’s actual permission set. The main exposure is not just data access, but persistent access that can outlive the original business need, especially when tokens, consent grants, or integration users are left in place after the app is no longer actively reviewed.

Failure mechanism: An attacker, compromised vendor account, or over-permissioned integration can abuse an OAuth grant or API token to read, export, or manipulate Salesforce data without triggering normal password controls. If the app was approved broadly, the blast radius can extend far beyond the original use case.

Impact: The result can be CRM data theft, fraudulent automation, exposure of customer or support records, and difficult-to-detect persistence inside a trusted business platform. In mature environments, the most damaging failures are usually governance failures first, then technical failures second.

Standards & Framework Alignment

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

CIS Controls v8 and NIST SP 800-53 Rev 5 set the technical controls, while ISO/IEC 27001:2022 defines the regulatory obligations.

FrameworkControl / ReferenceRelevance
CIS Controls v8CIS-5 — Account ManagementConnected app governance depends on knowing and removing active access paths.
Recommendation — Inventory and remove unused app access paths before they become standing exposure.
NIST SP 800-53 Rev 5AC-2 — Account ManagementIntegration users and app-linked accounts need ownership and lifecycle control.
AC-6 — Least PrivilegeSalesforce app scopes should be narrowed to the minimum required permissions.
AU-6 — Audit Record Review, Analysis, and ReportingUnusual IP, geolocation, and user-agent changes require routine log review.
Recommendation — Review and disable dormant integration accounts and app-linked access. Constrain each connected app to the minimum permissions its use case needs. Monitor app activity logs for anomalies that suggest token misuse or compromise.
ISO/IEC 27001:2022A.5.15 — Access ControlThird-party app access is an access-control decision over business data and functions.
A.8.2 — Privileged access rightsSome integrations effectively hold privileged access and need tighter governance.
Recommendation — Define approval and review rules for every external app connection. Treat powerful integrations as privileged access and review them more often.

Practitioner Guidance

What to prioritise: Start with connected apps that have broad scopes, long-lived tokens, no named business owner, or no recent business justification. Those are the highest-value removals because they combine weak governance with the largest potential blast radius.

What to verify: Confirm that every approved app has an owner, a current purpose, a documented permission scope, and a revocation path. If you cannot quickly answer who approved it, why it exists, and what data it can reach, the control is not mature enough to trust.

Practitioner takeaway: Salesforce app governance is strongest when approval, least privilege, and post-approval monitoring are treated as one control loop, not three separate tasks.

Free weekly newsletter

Subscribe to the NHI & AI Identity Journal

The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.

Bonus 33% off our NHI Course when you subscribe.

NHIMG Editorial Note
Reviewed and updated by the NHIMG editorial team on October 30, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org