TL;DR: Data discovery and classification tools now matter less for labeling alone than for linking sensitive data to identity, permissions, and remediation, according to Netwrix, because 26.4% of files uploaded to GenAI tools contained sensitive data and 46% of respondents experienced account compromise in 2025. Discovery without access context leaves risk in place.
At a glance
What this is: This is a Netwrix analysis of data discovery and classification tools that argues identity and permissions context determines whether discovery creates actual risk reduction or just better visibility.
Why it matters: IAM, IGA, and security teams need classification outputs that map to who can reach data, because unmanaged access turns discovery into reporting rather than governance.
Context
Data discovery and classification tools find sensitive information across repositories, but discovery on its own does not reduce exposure. The governance problem starts when teams can label data but cannot connect those labels to identity, effective permissions, or remediation paths.
In practice, that means the control boundary has to extend beyond content inspection. For IAM and IGA teams, the question is not only where sensitive data lives, but which users, groups, service accounts, and delegated identities can actually reach it and whether that access is still justified.
Netwrix frames the market around that distinction: some tools stop at classification, while others use identity context to turn discovery into action. That distinction is now central for hybrid environments, cloud estates, and data governance programmes dealing with GenAI exposure.
Key questions
Q: How should security teams connect asset discovery to identity governance?
A: Security teams should treat asset discovery as an input to identity governance, not as a separate inventory exercise. The right workflow links application presence, ownership, usage, and revocation evidence so access reviews, offboarding, and licence rightsizing can use the same data set. Without that linkage, the inventory is informative but not actionable.
Q: Why does classification without permissions context leave risk in place?
A: Because a sensitive file that remains broadly reachable is still exposed, even if it is perfectly labeled. Without permissions context, teams cannot tell whether access is stale, inherited, or appropriate. That means the organisation knows what is sensitive but not whether the exposure is actually controlled.
Q: What are the signs that discovery tools are only producing visibility, not control?
A: The main signs are reports that list sensitive repositories but do not trigger owner review, entitlement changes, quarantine, or enforcement actions. If the output stops at labels, the programme is still descriptive. Control exists only when the finding changes access, monitoring, or downstream policy.
Q: Should organisations prioritise access review or data labeling first?
A: Organisations should avoid treating them as separate tracks. Labeling helps identify sensitive data, but access review determines whether the data is actually exposed. If you can only do one first, start with the highest-risk repositories where sensitive data and broad access overlap, then use labels to scale the review.
Technical breakdown
Why identity context changes the value of classification
Data classification assigns sensitivity labels, but labels alone do not change exposure. Identity context adds the missing layer by resolving effective access, inherited permissions, and group membership so teams can see whether a sensitive file is actually reachable. In hybrid estates, this matters because the same dataset may be reachable through direct rights, nested groups, or stale permissions that are invisible in a content-only scan. Without that second layer, organisations produce inventories that look complete but do not support control decisions.
Practical implication: Treat classification as input to access analysis, not as the control outcome.
Why remediation is the real differentiator
A discovery tool that only reports sensitive files leaves the risk in place. The operational value appears when classification results trigger owner review, permission revocation, quarantine, or policy propagation into DLP and auditing systems. That shifts the tool from a visibility layer to a governance workflow. In practice, the best systems tie data findings to the identities and entitlements that created the exposure, then push that context into the processes that can fix it.
Practical implication: Prioritise platforms that can drive entitlement changes from classification results.
How GenAI exposure makes access governance harder
Data flowing into GenAI tools creates a new path for sensitive information to escape its original repository context. Once content is uploaded, copied, or shared into AI workflows, classification still matters, but the identity question becomes more urgent because access may now involve human users, service accounts, and tool integrations across multiple systems. That is why discovery must extend to where the data is going, not only where it started. The issue is not just leakage, but whether governance can keep up with data movement across toolchains.
Practical implication: Extend discovery and access reviews into AI and collaboration workflows, not only static storage.
Threat narrative
Attacker objective: Obtain or retain access to sensitive data by exploiting the gap between discovery and governed remediation.
- Entry occurs when sensitive data is uploaded, shared, or exposed in repositories that discovery tools later scan.
- Credential or permission exposure follows when overbroad identities, stale groups, or misconfigured rights make the data reachable.
- Escalation happens when classification lacks enforcement, so visible risk is not translated into revoked access, quarantine, or monitoring.
- Impact is persistent exposure of regulated or high-value data despite having identified it in the first place.
Breaches seen in the wild
- Azure Key Vault Contributor escalation 2024: Datadog found Azure Key Vault Contributor could add itself to access policies and read every secret, key and certificate in a vault.
- Microsoft SAS token exposure 2023: An over-permissive Azure SAS token in a Microsoft AI GitHub repo exposed 38TB, including workstation backups and Teams messages, for 3 years.
Read and download The State of NHI & AI Agent Breach Report 2026, covering 150+ breaches impacting Non-Human Identities including AI Agents.
NHI Mgmt Group analysis
Discovery without identity context is not governance: Classification tells you what the data is, but not who can reach it, how they got there, or whether that access still makes sense. That is why content-only tooling produces visibility without control. The practitioner lesson is that discovery programmes must be evaluated on whether they change entitlement decisions, not just reporting volume.
Identity-anchored classification is the point where data security becomes actionable: When sensitive data is mapped to effective access, stale groups, inherited rights, and mis-scoped permissions become visible as governance defects rather than abstract findings. That makes the access review process materially stronger because it is anchored in actual exposure, not theoretical sensitivity. The implication for IAM and IGA teams is that data context should feed entitlement workflows, not sit beside them.
Remediation, not labeling, is the control boundary: A labelled file that remains broadly reachable is still a live exposure. The market is moving toward tools that close the loop from discovery to owner review, revocation, quarantine, and audit evidence. Practitioners should treat any platform that stops at classification as incomplete for risk reduction.
Identity blast radius is the better metric for data discovery programmes: The real question is how many identities, groups, and delegated access paths can touch a sensitive dataset after it is found. That is a more useful measure than repository count because it ties data visibility to exposure scope. Teams should use that lens to prioritise cleanup where access radius is largest.
GenAI makes the governance gap more visible, not less: Sensitive content moving into AI workflows amplifies the weakness of discovery-only approaches because data can cross repository boundaries faster than access governance can adapt. This is where the overlap between data discovery, IAM, and shadow AI governance becomes operational. Practitioners need one view of data, identity, and downstream use or they will keep rediscovering the same risk in different places.
From our research library:
- Only 5.7% of organisations have full visibility into their service accounts, according to the Ultimate Guide to NHIs.
- Read next: NHI Lifecycle Management Guide
What this signals
Identity context is what turns data discovery from inventory into governance: Teams that can only classify sensitive content still lack the information needed to reduce blast radius. When effective access is tied to each finding, the programme can prioritise the identities and groups that create the most exposure, not just the repositories with the most labels.
Discovery programmes now need a direct path into entitlement cleanup: The practical test is whether a flagged dataset can drive owner review, permission revocation, or quarantine without a manual handoff. If that path does not exist, the organisation has visibility but not control, and the same access issue will reappear in the next scan.
For practitioners
- Map discovery findings to effective access Require every sensitive-data finding to show the identities, groups, and inherited rights that can reach it before it enters a remediation queue.
- Trigger entitlement reviews from classification output Use owner review and access certification workflows when classification identifies overexposed or stale repositories, rather than treating the finding as a standalone report.
- Extend governance into GenAI upload paths Include collaboration tools and generative AI upload points in discovery scope so exported sensitive data is assessed against the identities and integrations that can move it.
- Prioritise cleanup by identity blast radius Rank sensitive repositories by how many users, groups, service accounts, and delegated identities can access them, then reduce the widest exposure first.
- Verify remediation closes the loop Confirm that quarantine, permission revocation, label propagation, or DLP enforcement is actually triggered from the discovery workflow and not handled manually later.
Key takeaways
- Discovery tools are only partially useful when they stop at classification, because exposure depends on who can reach the data, not just where it resides.
- Identity context makes sensitive-data findings actionable by linking them to effective permissions, inherited rights, and access review workflows.
- The strongest programmes use discovery to drive remediation, access cleanup, and governance decisions across hybrid and AI-related data paths.
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 and MITRE ATT&CK address the attack and risk surface, while NIST CSF 2.0 and CIS Controls v8 set the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| OWASP Non-Human Identity Top 10 | NHI-05 — Overprivileged NHI | Overexposed datasets remain risky when identities have broader access than the data requires. |
| NHI-03 — Vulnerable Third-Party NHI | The article's access problem extends to delegated and integrated identities touching data paths. | |
| Recommendation — Review sensitive repositories for overprivileged access and reduce unnecessary reach. Inventory third-party and delegated identities that can reach sensitive data and revoke unused access. | ||
| NIST CSF 2.0 | PR.AA-05 — Access Permissions, Entitlements and Authorizations | The core issue is whether data findings map to real permissions and authorizations. |
| Recommendation — Link discovery results to entitlement reviews and remove unjustified access paths. | ||
| CIS Controls v8 | CIS-5 — Account Management | Identity context depends on knowing which accounts and groups can access sensitive data. |
| Recommendation — Maintain accurate account and group inventories so discovery findings can be tied to active access. | ||
| MITRE ATT&CK | TA0006;TA0008 — Credential Access; Lateral Movement | Sensitive-data exposure grows when excessive access enables movement across repositories and services. |
| Recommendation — Map exposed data paths to credential access and lateral movement risk in your detections. | ||
Key terms
- Identity context: The entitlement, ownership, and purpose information that explains why an action occurred and whether it was expected. For security operations, identity context turns raw alerts into decisions by showing which human or non-human identity acted and what it was allowed to do.
- Effective Permissions: Effective permissions are the access an identity can actually use after role inheritance, scope, and policy are applied. In Azure AI environments, they often matter more than the assigned role name because inherited rights can widen access to data, logs, and secret stores.
- Identity Blast Radius: The amount of damage a compromised identity can cause across systems, data, and infrastructure. In NHI environments, it is shaped by permissions, network reach, and administrative capability rather than by the credential alone. Reducing blast radius is a containment strategy that limits lateral movement and data exposure.
- Downstream Enforcement: The point at which identity decisions are actually applied by the systems that host data or services. A central source of truth only provides real control when those downstream systems receive and enforce changes such as revocation, role updates, and entitlement removal.
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 or identity security programme, it is worth exploring.
Published by the NHIMG editorial team on June 20, 2026.
Updated on October 8, 2026.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org