TL;DR: Identity resolution creates governance risk when linkage decisions ignore consent, jurisdiction, and purpose limitations, according to Transcend. Permission state has to follow reused identifiers across CRMs, warehouses, analytics, and AI pipelines or deletions and opt-outs will stay incomplete.
Editorial analysis by NHI Mgmt Group, based on content published by Transcend: “Identity resolution for enterprises: A privacy-safe guide”.
Key questions
Q: Why do deletion and opt-out requests fail in fragmented enterprise data estates?
A: They fail when the request is recorded in one system but not propagated to every downstream system that reuses the person's identifiers.
Q: When should privacy teams prioritise identity governance over broader data cleanup?
A: When rights requests, AI training, or customer profiling depend on knowing exactly which records belong to which person.
Q: How should organisations govern digital identity when AI is part of the service model?
A: They should treat identity assurance, access control, and AI oversight as one governance chain.
Practitioner guidance
- Define identity resolution as a governed workflow Document which identifiers may be linked, under what consent basis, and in which jurisdictions the linkage is permitted.
- Propagate opt-outs and deletions across every system Test whether preference changes, suppression flags, and erasure requests reach CDPs, CRMs, warehouses, SaaS tools, and AI pipelines without manual reconciliation.
- Audit identity mappings for stale or conflicting records Look for duplicate profiles, outdated joins, and unresolved identifiers that can cause false matches or incomplete rights handling.
Bottom line: Identity resolution creates governance risk when linkage decisions ignore consent, jurisdiction, and purpose limitations.
Glossary terms: Identity Resolution, Repository Propagation, Identity Graph, Data Flow Governance
Explore further
View Full Forum → | NHI Foundation Course → | Our Services → | Read the full analysis →
Identity resolution is now a governance control, not a matching problem. The central issue is not whether organisations can connect records, but whether they can justify every link they make. Once a person's identifiers are spread across warehouses, CRM platforms, support tools, and AI pipelines, the matching layer becomes a policy boundary. Practitioners should treat linkage rules as part of data governance architecture, not as a back-office data quality task.
A question worth separating out:
Q: What controls help prevent AI models from training on unapproved personal data?
A: Require permission state to travel with the data lineage from source to training set, then enforce suppression at every reuse point. That means approvals, deletions, and opt-outs must be visible to data science workflows, not just to privacy operations. If the model can access stale or unresolved records, governance has already failed.
👉 Read our full editorial: Identity resolution needs governance, not just better matching