Sector guidesUpdated 22 November 20269 min read

DPDP for Hospitals, Clinics and Health-Tech

How hospitals, clinics, diagnostic labs and health-tech apps handle consent, medical-emergency processing, children's records, breach and AI under the DPDP Act and Rules.

By Suhas Bhairav · Reviewed against the DPDP Rules, 2025 (G.S.R. 846(E), 13 Nov 2025)

Applies from
13 May 2027
Emergency processing
Section 7 legitimate uses (medical emergency, epidemic)
Children's exemption
Fourth Schedule Part A items 1–2
Breach reporting
Every breach; 72-hour detailed Board report
Log retention
Minimum one year (Rule 8(3))
Security penalty
Up to ₹250 crore

Quick answer

Hospitals, clinics and health-tech apps are Data Fiduciaries under DPDP from 13 May 2027. Routine care typically relies on consent or a legitimate use, with section 7's medical-emergency and epidemic uses covering urgent cases. The Fourth Schedule exempts clinical establishments and healthcare professionals from parental-consent and tracking rules only for health services to a child, to the extent necessary.

  • DPDP has no separate 'sensitive' category, but health data's sensitivity weighs in SDF designation and penalty decisions.
  • Section 7 permits processing without consent for medical emergencies involving threat to life and for health services during epidemics.
  • Fourth Schedule Part A exempts clinical and mental health establishments and healthcare professionals from s. 9(1) and 9(3) for children's health services.
  • Allied healthcare professionals are exempt only for supporting a recommended treatment and referral plan.
  • Every breach of patient data must be reported, with a detailed Board report within 72 hours.

Who this is for: Hospitals and clinic chains, Diagnostic labs and pharmacies, Telemedicine and health-tech apps, Mental health providers, Health AI and clinical decision support vendors

What lawful basis do hospitals and health apps use under DPDP?

A hospital, clinic, lab or health app that decides why and how patient data is processed is a Data Fiduciary. The DPDP Act gives two routes:

  • Consent (section 6) — free, specific, informed, unconditional, unambiguous, with clear affirmative action, limited to necessary data, and withdrawable as easily as given. Planned treatment, teleconsultations, health-app accounts, wearables and research participation typically rely on consent with a Rule 3 notice.
  • Legitimate uses (section 7) — including data voluntarily provided for a specified purpose where the patient has not indicated non-consent (section 7(a)); responding to a medical emergency involving a threat to life or immediate threat to health; and providing medical treatment or health services during an epidemic, outbreak or other threat to public health.

The emergency and epidemic uses cover situations where asking for consent is impractical. They are not a general basis for routine care, marketing or analytics.

DPDP has no separate "sensitive personal data" category — unlike the SPDI Rules, 2011, which apply until 13 May 2027 when section 43A of the IT Act is omitted by section 44(2) of the DPDP Act. Health data's sensitivity still matters: it is a factor in designating Significant Data Fiduciaries and in the Board's penalty decisions.

Health-sector laws and professional rules on medical records that are stricter continue to apply alongside DPDP.

Children's health data: the Fourth Schedule exemption

Section 9(1) requires verifiable parental consent before processing a child's data, and section 9(3) bars tracking and behavioural monitoring of children. Rule 12 and Fourth Schedule Part A carve out healthcare:

Fiduciary classCondition
Clinical establishment, mental health establishment or healthcare professional (item 1)Processing restricted to provision of health services to the child, to the extent necessary for protection of her health
Allied healthcare professional (item 2)Processing restricted to supporting implementation of a healthcare treatment and referral plan recommended by such professional, to the extent necessary for protection of her health

The Schedule's note takes definitions from other laws: "clinical establishment" from the Clinical Establishments (Registration and Regulation) Act, 2010; "mental health establishment" from the Mental Healthcare Act, 2017; "healthcare professional", "allied healthcare professional" and "health services" from the National Commission for Allied and Healthcare Professions Act, 2021.

Practical limits:

  • The exemption covers health services, not paediatric marketing, app engagement analytics or targeted ads.
  • A health-tech app that is not itself a clinical establishment or professional should not assume item 1 applies; it may act as a processor for one.
  • Section 9(2) — no detrimental processing — is not exempted.

See children's data exemptions.

Security and breach for clinical systems

Hospital systems — HIS, EMR, PACS imaging, LIS, pharmacy, billing, insurance desks — are often old, interconnected and widely accessed. Rule 6 sets minimum safeguards from 13 May 2027:

  • Encryption, masking or tokens (6(1)(a)). Encrypt EMR databases and imaging archives; mask patient identifiers in analytics and research extracts.
  • Access control (6(1)(b)). Role-based access for clinicians, nurses, billing and IT; break-glass for emergency access with mandatory review.
  • Logs and monitoring (6(1)(c)). Log every record access; review for snooping on VIP or staff records.
  • Continuity (6(1)(d)). Tested backups and recovery for ransomware scenarios, which threaten availability as well as confidentiality.
  • Log retention (6(1)(e)). Keep access logs and relevant data for one year.
  • Vendor contracts (6(1)(f)). HIS vendors, cloud PACS, transcription and AI vendors.

Rule 7 makes every personal data breach reportable — no threshold. Notify each affected patient without delay through registered contact (for a child, the parent), and the Board immediately with a description, then a detailed report within 72 hours. CERT-In requires specified incidents to be reported within 6 hours. Ransomware that encrypts patient records is a breach event even if no data is exfiltrated, because availability is compromised. See the breach playbook.

Health AI: scribes, triage and clinical decision support

AI is arriving in clinics through ambient scribes, triage chatbots, radiology assistance and coding automation. Each adds data flows.

  • Ambient scribes. Consultations are recorded and transcribed. Tell patients, obtain consent where you rely on it, and set short retention for raw audio once the note is signed — subject to the Rule 8(3) one-year minimum for processing logs and any medical-records law.
  • Triage chatbots. Patients disclose symptoms before any clinician relationship exists. Use a Rule 3 notice up front and route emergencies to humans; the medical-emergency legitimate use helps only where there is a genuine threat to life or health.
  • Imaging and diagnostics AI. Images often carry identifiers in metadata. Strip or tokenise before sending to vendors.
  • Model vendors. Contractually prohibit training on patient data unless specifically agreed and lawful.
  • Research. Rule 16 exempts processing necessary for research, archiving or statistical purposes only when done per Second Schedule standards — lawful, purpose-limited, necessary, accurate, retention-limited, secure and accountable.

Mask identifiers before LLM calls and keep a gateway log. See masking for AI pipelines and voice AI and call recordings.

Patient rights, records and retention

Patients are Data Principals with rights under sections 11–14 from 13 May 2027:

  • Access to a summary of personal data processed and processing activities, and identities of fiduciaries and processors it was shared with (labs, insurers, TPAs).
  • Correction, completion and updating; erasure where the purpose is served and no law requires retention.
  • Grievance redressal, with a published response period of not more than 90 days (Rule 14(3)).
  • Nomination of individuals to exercise rights on death or incapacity (Rule 14(4)).

Rule 14(1) requires publishing how to make requests and which identifiers are needed — for hospitals, typically UHID or patient ID, registered mobile and date of birth.

Retention is where health and DPDP rules intersect. Medical-records retention requirements under other laws and professional regulations generally require clinical records to be kept for defined periods; those prevail. Erasure requests for clinical records will often be answered by explaining the legal retention. Non-clinical data — marketing lists, app telemetry, chatbot logs — should follow DPDP purpose limits, with Rule 8(3)'s one-year minimum for logs.

Publish the Rule 9 contact on your website and include it in every response to a rights request.

Hospital and health-tech readiness plan

  1. Inventory clinical, administrative, marketing and AI systems, and every vendor that touches patient data.
  2. Basis per purpose — consent, section 7(a), emergency, epidemic — documented by department.
  3. Notices under Rule 3 at registration desks, on apps and for teleconsultation, in languages patients use.
  4. Access control cleanup — remove shared logins, enforce break-glass, review access logs.
  5. Children's processing — confirm Fourth Schedule item 1 or 2 applies, and that anything outside health services gets parental consent.
  6. AI controls — scribe consent, masking, vendor no-training clauses.
  7. Breach drills — ransomware tabletop covering 6-hour CERT-In reporting and 72-hour Board report.
  8. Rights desk — published channels, identifiers, 90-day maximum response, Rule 9 contact.
  9. Retention schedule reconciling medical-records law with DPDP and Rule 8(3).
  10. Staff training for front-desk, nursing and IT teams on recognising a breach and escalating it within hours.

Large hospital networks and health platforms should also prepare for possible SDF notification, given data sensitivity.

Further reading on suhasbhairav.com

Before 13 May 2027

Compliance checklist

  1. Document the lawful basis per clinical and non-clinical purpose

    Map consent, s. 7(a), medical-emergency and epidemic uses to each data flow.

    DPDP Act ss. 6–7

  2. Confirm the Fourth Schedule healthcare exemption scope

    Rely on Part A items 1–2 only for health services or treatment plans to the extent necessary.

    Rule 12; Fourth Schedule Part A

  3. Enforce role-based EMR access with break-glass

    Limit record access by role and review every emergency override.

    Rule 6(1)(b)

  4. Log and review record access

    Monitor for unauthorised viewing and keep logs for one year.

    Rule 6(1)(c), 6(1)(e)

  5. Test backups against ransomware

    Prove recovery of EMR, imaging and lab systems.

    Rule 6(1)(d)

  6. Bind HIS, cloud and AI vendors by contract

    Include safeguards, breach notice, retention and no-training terms.

    Rule 6(1)(f)

  7. Get consent for ambient AI scribes

    Inform patients and record consent where relied upon; set audio retention.

    DPDP Act s. 6; Rule 3

  8. Prepare 72-hour breach reporting

    Notify affected patients without delay and the Board with a detailed report within 72 hours.

    Rule 7

  9. Publish patient rights channels

    List identifiers such as patient ID and set a grievance response of at most 90 days.

    Rule 14

  10. Use Rule 16 only for genuine research

    Follow Second Schedule standards for research and statistical processing.

    Rule 16

Key dates

  1. DPDP Rules notified

  2. Core DPDP obligations apply; IT Act s. 43A and SPDI Rules cease

Answers

Frequently asked questions

Do hospitals need patient consent under DPDP?

For most planned care, apps and teleconsultations, consent or the section 7(a) voluntary-provision legitimate use is the basis. Section 7 also allows processing without consent to respond to a medical emergency involving threat to life or health, and for health services during an epidemic or public health threat.

Is health data sensitive personal data under the DPDP Act?

The DPDP Act has no separate sensitive personal data category, unlike the SPDI Rules, 2011, which cease with the omission of section 43A of the IT Act on 13 May 2027. Health data's sensitivity still matters for Significant Data Fiduciary designation and for the Board's assessment of penalties.

Do paediatric clinics need parental consent under DPDP?

Fourth Schedule Part A item 1 disapplies sections 9(1) and 9(3) for clinical establishments, mental health establishments and healthcare professionals, but only for providing health services to the child to the extent necessary to protect her health. Processing beyond that, such as marketing, still needs verifiable parental consent.

Is ransomware on a hospital system a DPDP breach?

Yes. A personal data breach under the Act includes compromise of the availability of personal data, not only theft. Rule 7 requires every breach to be notified to affected patients without delay and to the Board, with a detailed report within 72 hours. CERT-In's 6-hour reporting also applies.

Can a patient ask a hospital to delete medical records?

She can request erasure under section 12, but section 8(7) allows retention where it is necessary to comply with law. Medical-records retention requirements generally keep clinical records for defined periods. The hospital should explain the legal retention in its response, and erase non-clinical data that is no longer needed.

Can hospitals use patient data to train AI models?

Only with a lawful basis for that purpose. Treatment consent does not automatically extend to model training. Options include specific consent, robust de-identification, or research under Rule 16 following Second Schedule standards. Contracts with AI vendors should bar training on patient data unless lawful and agreed.

Legal references and sources

  • Digital Personal Data Protection Act, 2023 — Sections 6 and 7 (official text)
  • Digital Personal Data Protection Rules, 2025 — Rule 12; Fourth Schedule Part A items 1–2 (official text)
  • Digital Personal Data Protection Rules, 2025 — Rules 6 and 7 (official text)
  • Digital Personal Data Protection Rules, 2025 — Rule 16 and Second Schedule (official text)

This guide explains the law for engineering and business teams. It is not legal advice; confirm how it applies to your organisation with qualified counsel, and check the Gazette for later notifications or amendments.

About the author

Suhas Bhairav

Suhas Bhairav

Systems architect and applied AI practitioner who designs production AI systems for Indian and global companies, with a focus on privacy-by-design under the DPDP Act and the EU AI Act.