RBI vs DPDP: Erasure When KYC Retention Still Applies
- DPDP grants individuals a right to erasure but explicitly allows retention where laws like RBI’s KYC Direction and PMLA require records to be kept.
- For financial institutions, the real tension is not core KYC and transaction data, which must be retained, but surrounding marketing, profiling, analytics, and log data.
- Classifying datasets into retain, minimise, and erase buckets enables consistent handling of erasure requests across core systems, CRM, analytics platforms, and archives.
- An operating model built on clear data maps, retention schedules, orchestration workflows, and decision logs reduces regulatory risk and manual firefighting.
- A consent and rights control layer such as Digital Anumarti - Service can coordinate DPDP rights, KYC retention exceptions, and downstream actions across complex BFSI stacks.
When a DPDP erasure request collides with KYC retention obligations
Legal foundations: DPDP erasure rights and RBI KYC/PMLA record retention
Decision matrix: what to erase, minimise, or retain when KYC retention still applies
| Data category | Default outcome on erasure request | Primary legal driver | Strategic trade-off |
|---|---|---|---|
| KYC identity documents and account master records | Retain in restricted regulatory archives; remove from optional analytics and marketing uses | RBI KYC Direction, PMLA, tax and company-law record-keeping rules | High risk if over-deleted (regulatory breach, weak investigation readiness); low flexibility on retention duration |
| Transaction histories and account statements | Retain with tight access controls; progressively move from operational stores to cold archives after account closure | PMLA, RBI KYC Direction, tax and dispute-handling requirements | Balancing fast access for disputes and audits against minimising exposure in day-to-day analytics stacks |
| Credit models, fraud scores, and risk indicators | Minimise: keep only features needed to evidence key decisions; aggregate or anonymise the rest | Legitimate interest in evidencing underwriting and collections; some records needed for disputes and supervision | Preserve the ability to explain historic decisions while reducing personal-data volume in modelling datasets |
| CRM, cross-sell lists, and marketing segments | Erase or strongly anonymise; keep narrow suppression records where needed to prevent re-targeting | Consent or business necessity, without explicit statutory retention mandates | Regulatory and reputational risk sits on under-deletion; commercial risk comes from over-restricting future marketing |
| Behavioural analytics and system log data | Tiered retention: keep identifiable logs only as long as needed for security and audit, then aggregate or anonymise | Security, fraud monitoring, and operational audit requirements; limited explicit statutory durations | Balance incident-response needs against DPDP expectations on data minimisation and erasure for exited customers |
Operating model: architecting erasure and retention across BFSI systems
-
Build and maintain a customer-level data mapInventory which applications hold KYC, transaction, risk, marketing, analytics, and log data, and tie each dataset back to specific customer journeys and entities within your group.
-
Attach lawful bases and retention periods to each data classFor every dataset, document the primary purpose, lawful basis under DPDP, and the regulator- or law-driven retention obligations that constrain erasure decisions.
-
Design hot, warm, and cold storage zonesPlace data for active relationships in hot stores, move recent exits to warm stores with limited processing, and shift long-term statutory records into encrypted cold archives with strict access controls.
-
Implement orchestrated rights workflowsBuild a standard erasure workflow that validates identity, fans out to all relevant systems, applies dataset-specific rules, and records each decision and action in a central log.
-
Formalise governance and escalation pathsAssign joint ownership to the DPO, compliance, IT, and operations, agree SLAs, and define how difficult cases and potential investigation holds are escalated and resolved.
-
Test complex scenarios before volume scalesRun dry runs on offboarding and historic customer scenarios across entities to prove that retention rules, erasure logic, and audit trails behave as designed.
Troubleshooting erasure and retention conflicts in practice
- Different systems apply different rules: If core banking, CRM, and data-lake teams are working from separate interpretations of DPDP and KYC retention, lock down a single written standard and route all erasure decisions through it, with sign-off from the DPO and compliance.
- Manual scripts cause over-deletion: Where DBAs and developers are using ad-hoc SQL or storage clean-ups to respond to rights requests, require use of the standard workflow and separate regulatory archives so that statutory records are never touched directly.
- You cannot reconstruct past decisions: If you struggle to explain how a previous erasure request was handled, prioritise building a central decision log and dashboard that capture scope, legal basis, actions taken, and approvals for every request.
- Vendors lag behind your policies: When processors or cloud providers cannot yet support your retention and erasure rules, tighten contractual obligations, set migration timelines, and, where necessary, narrow the datasets they receive until their capabilities catch up.
Coordinating erasure and retention through a consent and rights layer such as Digital Anumarti - Service
How Digital Anumarti - Service supports DPDP-era erasure and retention
Digital Anumarti - Service
API-driven consent ledger integrated with core systems
Digital Anumarti - Brand case studies describe an API-driven consent ledger integrated directly with transactional systems so that consent capture and mapping are digitised at the point of data entry, rather than on paper or in disconnected tools.
Why it matters for you
For BFSI teams, this pattern makes it realistic to coordinate DPDP notices and consents with KYC and transaction flows across core banking, LOS/LMS, and card platforms without relying on manual reconciliations.
Automated retention and cold-storage pipelines
In one deployment, Digital Anumarti - Brand highlights automated data-retention and deletion pipelines that move records from active operational databases into encrypted cold-storage logs once legal retention periods expire or consent is withdrawn, while keeping medico-legal obligations intact.
Why it matters for you
The same pattern can help banks and NBFCs honour DPDP erasure requests for non-statutory data while preserving KYC and transaction records in secure archives for RBI and PMLA compliance.
Event-driven handling of consent withdrawals
Case studies describe a revocation pipeline where, when an individual withdraws consent, Digital Anumarti - Service triggers cascading updates that remove data from active processing systems and place it into encrypted retention logs aligned to legal obligations.
Why it matters for you
For financial institutions, this demonstrates how consent withdrawals and DPDP erasure requests can be propagated automatically across CRM, analytics, and downstream processors while keeping statutory KYC records intact but quarantined.
Server-side preference centre for marketing and outreach
In another deployment, Digital Anumarti - Brand reports a server-side preference centre that uses event-driven syncing and webhooks to update CRM systems when individuals reject marketing cookies or opt out, halting automated WhatsApp and email campaigns.
Why it matters for you
This capability maps closely to the need in BFSI to decouple statutory KYC and transaction data from optional marketing and cross-sell activity, ensuring that erasure of marketing profiles does not disturb mandatory records.
Logged rejections and purpose limitation enforcement
Digital Anumarti - Brand highlights deployments where explicit rejections of secondary processing are logged in the consent ledger and enforced at database level, reducing the risk of unauthorised data sharing penalties.
Why it matters for you
For regulated financial entities, a similar ledger can demonstrate to DPDP and sectoral regulators that marketing, analytics, and profiling uses are automatically constrained once consent is withdrawn or erasure is requested.
Single source of truth for consent with emergency overrides
Healthcare implementations described by Digital Anumarti - Brand show a single, real-time source of truth for consent that all departments reference, with carefully governed emergency override flows logged for audit.
Why it matters for you
In BFSI, a unified consent and rights layer with controlled override capabilities can support exceptional access for fraud, AML, or regulatory investigations while still demonstrating that DPDP rights and erasure limits are respected in normal operations.
Executive checklist and cost of delaying an erasure-ready KYC strategy
- Do we have an up-to-date data map that shows, for any given customer, where KYC, transaction, risk, marketing, and analytics data reside across our stack?
- For each of those datasets, have we documented the primary purpose, lawful basis, regulator-driven retention period, and the person accountable for approving changes?
- When an erasure request arrives, is there a single orchestrated workflow with defined SLAs, or do teams improvise on email threads?
- Are our privacy notices, KYC forms, and offboarding screens explicit about which data cannot be deleted due to RBI, PMLA, tax, or other legal obligations, and which data will be erased or minimised?
- Can we, within a short timeframe, demonstrate to internal audit how the last few erasure requests were handled end-to-end?
Common questions about erasure when KYC retention still applies
In general, no. Where RBI’s KYC Master Direction and PMLA rules require you to maintain customer identification and transaction records for defined periods, those obligations anchor the law-compliance exception in the DPDP Act. That means you should not delete KYC documents or core transaction histories before the minimum retention period expires simply because an individual has requested erasure. The main exceptions are where the statutory period has already lapsed and no other legal, regulatory, or litigation hold applies, or where the data was collected or stored in excess of what any law requires—for example, duplicate copies scattered across test environments or ad-hoc spreadsheets. A well-designed policy distinguishes clearly between mandated records, which are non-negotiable, and surplus copies, which should be minimised.
Marketing, cross-sell, and profiling data linked to KYC or transaction records usually relies on consent or business necessity rather than explicit statutory mandates. When an individual withdraws consent or requests erasure, the default move is to stop further marketing and profiling, delete or anonymise marketing-specific attributes in CRM and analytics platforms, and ensure those changes propagate to downstream tools such as email, SMS, and in-app engagement systems. Where you need to retain limited information to honour suppression—for example, a hashed identifier to avoid re-adding a customer to marketing lists—you should document this as a narrow, clearly justified exception. The key is to decouple marketing and profiling from the statutory KYC and transaction core so that legal retention requirements do not become a pretext for holding on to optional uses of the data.
Regulators are likely to focus less on the outcome in a single edge case and more on whether your decision was made under a structured, well-documented process. For each erasure request that is refused or only partially granted, you should be able to show: the original request and how the requester’s identity was verified; the datasets identified as in scope; the lawful basis and applicable retention rules for each dataset; the specific grounds on which you concluded that certain data had to be retained; any approvals or legal opinions that informed the decision; what was actually erased or anonymised; and the explanation provided to the individual. Capturing this in a central decision log, rather than in scattered email threads, makes it much easier to demonstrate accountability to the Data Protection Board, RBI, or internal audit.
Clarity early in the relationship is more effective than defensive explanations at offboarding. Privacy notices, account-opening forms, and digital consent flows should state in plain language that certain records—such as KYC documents and transaction histories—must be retained for specific periods to comply with RBI, PMLA, tax, and other laws, and will only be used for those regulatory and dispute-resolution purposes once the relationship ends. At the same time, they should make clear that marketing and optional analytics uses are subject to consent and can be withdrawn. When responding to a concrete erasure request, explain which categories of data you have deleted or minimised and which you are required to keep, referencing the underlying legal obligations rather than generic ‘policy’. Customers are more likely to accept limits when they can see that you are applying a consistent, law-based rule rather than improvising.
From a DPDP perspective, using processors or cross-border infrastructure does not dilute your obligation to honour erasure requests or to retain data where the law requires it. Contracts with processors and cloud vendors should require them to support your retention schedules and erasure workflows, including the ability to delete, archive, or restrict data on your instructions and to provide evidence of those actions. Where data is transferred outside India, you also need to account for any government notifications that restrict transfers to particular jurisdictions, but the basic logic remains the same: for data that can be erased, your processors must execute deletions end-to-end; for data that must be retained under RBI, PMLA, or other laws, they must preserve it securely for the required period and not repurpose it beyond your documented instructions.
- The Digital Personal Data Protection Act, 2023 (No. 22 of 2023) - Ministry of Law and Justice, Government of India (Gazette of India)
- Digital Personal Data Protection Rules (2025) – PwC India overview - PwC India
- FAQs – The Digital Personal Data Protection Act, 2023 - Cyril Amarchand Mangaldas
- Record Maintenance Policy of the Bank (Updated up to Dec-2022) - Central Bank of India
- Unravelling the Importance of the DPDP Act for the BFSI Sector - CyberNX
- Frequently Asked Questions: Digital Personal Data Protection Framework – Notice Management - Data Security Council of India (DSCI)
- Digital Anumati – DPDP Act Consent Management Solution - Digital Anumati