[Analysis] Data Fiduciary, Data Processor or Both – The Control Test | Role Classification Register | Processor Obligations
- Blog|Company Law|
- 15 Min Read
Determining and documenting the role under sections 2(i) and 2(k): the control test, joint fiduciary status under the phrase “in conjunction with other persons”, the five fields every role determination must record, the minimum that section 8(2) and Rule 6(1)(f) set for the processor contract, and the obligations a Data Processor actually bears.
- By Taxmann
- |
- Last Updated on 16 November, 2026
Table of Contents
- Why does the role matter?
- What is a Data Fiduciary?
- What is a Data Processor?
- Can an organisation be both?
- Where does the line fall in practice?
- Does the Act recognise joint fiduciaries?
- How should a role determination be documented?
- What does the law require of the processor contract?
- What obligations does a Data Processor actually bear?
- What does an auditor test, and what goes wrong?
- Quick answers
1. Why does the role matter?
Because the Act allocates obligations by role, and the allocation is uneven.
A Data Fiduciary bears the primary obligations under sections 4 through 10 and is subject to the penalty framework in the Schedule. A Data Processor has a more limited set of direct obligations but is subject to the security safeguard requirements and the contractual provisions mandated by section 8(2) and Rule 6(1)(f) (para 1.9.3).
The book sets out the practical cost of getting it wrong, and it runs in both directions. Applying the wrong classification to a processing activity means either assuming obligations that do not apply, which creates unnecessary compliance overhead, or failing to discharge obligations that do apply, which creates enforcement risk (para 2.14.1).
The first of those failures is more common than it sounds. A Data Processor that has been incorrectly classified as a Data Fiduciary in internal records will receive misdirected instructions: it will be asked to issue notices and collect consent for processing activities where the notice and consent obligations rest with the contracting Data Fiduciary (para 2.14.2). Two organisations each believing the other is issuing the notice is how a notice fails to be issued at all.
The determination is also not a one-time exercise. As an organisation’s operations evolve, as it launches new products, enters new markets, engages new vendors or changes its data architecture, the role classification must be revisited periodically (para 2.14.1).
2. What is a Data Fiduciary?
Section 2(i) defines a Data Fiduciary as any person who alone or in conjunction with other persons determines the purpose and means of processing of personal data, and the test is control: the question is not who physically handles the data, it is who decides why the data is processed and how (para 2.3). The definition, the book’s reading of it and the obligations that follow are set out in the meaning and obligations of a Data Fiduciary.
For the classification question, the illustration that matters is the book’s payroll case. An organisation that engages an external payroll provider to process employee salary data remains the Data Fiduciary in respect of that processing, because it determines the purpose, salary payment, and the means, the categories of data to be processed, the frequency of processing and the retention period. The payroll provider, which acts on the instructions of the Data Fiduciary, is the Data Processor (para 2.3).
The role is not confined to companies. “Person” is defined broadly under section 2(s), covering an individual, a Hindu undivided family, a company, a firm, an association of persons or body of individuals whether incorporated or not, the State, and every artificial juristic person (Glossary).
3. What is a Data Processor?
Section 2(k) defines a Data Processor as any person who processes personal data on behalf of a Data Fiduciary (para 2.4).
The defining characteristic is the “on behalf of” relationship. A Data Processor does not determine the purpose and means of processing; it executes instructions received from the Data Fiduciary. The Data Fiduciary decides what data to process, for what purpose and under what conditions. The Data Processor performs the processing activity (para 2.4).
The consequence is that the obligations which define the Act’s relationship with individuals sit with the Data Fiduciary. The notice obligation under section 5 falls on the Data Fiduciary. The consent obligation under section 6 falls on the Data Fiduciary. The obligation to ensure data quality under section 8(3) falls on the Data Fiduciary. The obligation to establish a grievance redressal mechanism under section 8(10) falls on the Data Fiduciary (para 2.4).
What the Data Processor carries is derivative but real, and section 9 of this article sets it out in full.
The accountability rule that governs the whole relationship is section 8(1), and the book quotes its effect directly. A Data Fiduciary shall be responsible for complying with the provisions of the Act and the Rules in respect of any processing undertaken by it or on its behalf by a Data Processor, irrespective of any agreement to the contrary. A contractual clause that purports to transfer the Data Fiduciary’s compliance obligations to the Data Processor does not relieve the Data Fiduciary of liability. The Data Fiduciary may delegate the processing. It cannot delegate the accountability (para 2.4).
4. Can an organisation be both?
Routinely, and this is where most classification exercises go wrong.
The book states the rule at the start: the exercise proceeds at the level of individual processing activities, not at the organisational level. An organisation may be a Data Fiduciary for its employee data processing, a Data Processor for client data it hosts under a service agreement, and a joint Data Fiduciary, in conjunction with a partner entity, for co-branded marketing activities. Each of those classifications carries a different set of obligations under the Act (para 2.14.1).
Its worked example is an IT services company. The company that processes employee personal data for its own HR purposes is a Data Fiduciary in respect of that processing. The same company, when it processes client customer data under a service agreement, is a Data Processor in respect of that processing. The role classification must be determined at the level of each processing activity, not at the organisational level (para 1.9.3).
This matters for how the exercise is scoped. A single organisational answer, recorded once in a policy, is not a role classification. What the Act’s obligation structure requires is a register with one row per processing activity.
5. Where does the line fall in practice?
The book identifies the hard case and declines to give a bright-line answer, which is the correct treatment because the Act does not provide one.
The practical question that arises in many engagements is where the line falls between a Data Processor and a Data Fiduciary. A cloud infrastructure provider that offers storage and compute resources, where the customer determines what data is stored and how it is processed, is ordinarily a Data Processor. But a cloud-based customer relationship management platform that determines the structure of data fields, the analytics algorithms applied to customer data and the retention policies may cross the line into Data Fiduciary territory for some processing activities. The classification must be assessed for each processing activity on its facts and purpose, not at the entity level (para 2.4).
Read against the section 2(i) test, the signal is specific: a provider that makes decisions about purpose or means, as distinct from executing the customer’s decisions, is determining purpose and means for that activity. Setting the data model, selecting the analytics applied, and fixing the retention period are all decisions about means. A provider that does all three while describing itself as a processor has a classification problem regardless of what the contract says.
Where the classification is contested or ambiguous, as may occur in complex outsourcing or platform arrangements, the book’s recommendation is that a legal opinion supporting the determination is advisable (para 2.14.2).
6. Does the Act recognise joint fiduciaries?
Not by that name, and in substance yes.
The phrase “alone or in conjunction with other persons” in section 2(i) accommodates the reality that processing activities frequently involve multiple entities making decisions about purpose and means. The DPDP Act does not use the term “joint controller” or “co-fiduciary”, but the book’s reading is that the phrase creates the same conceptual space. Where two or more entities jointly determine the purpose and means of processing, each entity is a Data Fiduciary in respect of that processing and bears the full set of obligations under the Act (para 2.3).
The book draws on Fashion ID GmbH & Co. KG v. Verbraucherzentrale NRW eV (CJEU, 2019) to show how far the concept reaches. The Court held that a website embedding the Facebook “Like” button is a joint controller with Facebook for the collection and transmission of visitors’ data, even though the site never sees the data and Facebook does the downstream processing. The decision established that joint controllership arises from the functional test of influence over purposes and means, not from physical access to the data. That principle applies by analogy to the DPDP Act’s concept of a Data Fiduciary determining purpose and means in conjunction with other persons (para 2.3).
Translated into practice, every embedded third-party pixel, tag, software development kit and social plugin on a website or in an application is a candidate for joint determination, because the organisation decides to deploy it and thereby influences the purpose and means of the collection it performs.
One question the Act leaves open, and the book says so, is how liability is allocated between joint Data Fiduciaries, whether each is liable for the whole of the processing or only for the portion it controls. That is not expressly addressed in the Act and will likely be resolved through Board proceedings and, eventually, appellate jurisprudence (para 2.3). Until it is, the prudent reading is that each joint Data Fiduciary should assume it bears the full set of obligations, and should allocate responsibility between the parties by contract while recognising that section 8(1) prevents the contract from displacing its own accountability.
7. How should a role determination be documented?
In a register, with five fields per processing activity. The book sets them out (para 2.14.2):
| Field | What to record |
|---|---|
| 1 | A description of the activity and the categories of personal data involved |
| 2 | The identity of the entity or entities that determine the purpose of processing |
| 3 | The identity of the entity or entities that determine the means of processing |
| 4 | The classification: Data Fiduciary, Data Processor, or joint Data Fiduciary |
| 5 | The date and authorship of the assessment |
Fields 2 and 3 are separate for a reason. Section 2(i) requires the determination of purpose and means, so splitting them is what makes the conclusion auditable rather than asserted. Field 5 is the one most often omitted, and it is the field that gives the record its evidentiary value.
The register itself should map each processing activity to the applicable role, the basis for the classification, meaning who determines the purpose and who determines the means, and the legal reference supporting the determination. The book describes it as a foundational document in the compliance programme, to be reviewed at defined intervals, annually at a minimum, and in immediate response to any material change in processing activities (para 2.14.1).
The register has to reach every engagement in which personal data is handled on the organisation’s instruction. In practice that means payroll providers, cloud and hosting platforms, customer-support and marketing tools, analytics providers, and the compliance software used to prepare statutory filings that carry employee and vendor personal data. Each of those is a row, with its own determination and its own contract position.
The book gives two reasons for documenting, and the second is the one that matters in a proceeding. First, internal clarity: business units, IT teams, procurement functions and legal teams need to know which obligations apply to which processing activities. Second, evidentiary support in the event of a Board inquiry. Section 6(10) places the burden of proving valid notice and consent on the Data Fiduciary, and if there is a dispute about whether an entity is a Data Fiduciary or a Data Processor in respect of a particular processing activity, the contemporaneous documentation of the role determination will carry significant weight (para 2.14.2).
Appendix A to the book lists the evidence this requirement calls for: a role classification register mapping each processing activity to the applicable role; an internal policy or memorandum documenting the role determination criteria; board or management approval of the role classification; and periodic review records of role assessments (Appendix A, section 1).
Role classification also sits in the first ninety days of the compliance programme. Figure 12.1, the DPDP Compliance Readiness Checklist, places “role classification completed as Data Fiduciary, Data Processor or both” in its Discovery block, alongside the personal data inventory, the data flow maps and the lawful basis register (para 12.1.3). The sequence is deliberate: the inventory tells you what processing exists, the flow maps tell you who touches it, and the role determination tells you which obligations attach.
8. What does the law require of the processor contract?
Two things, and the book is precise about both. Section 8(2) provides that a Data Fiduciary may engage, appoint, use or otherwise involve a Data Processor to process personal data on its behalf for any activity related to the offering of goods or services to Data Principals only under a valid contract. A Data Fiduciary that lets a vendor process personal data without a written agreement, whether through an informal arrangement, a verbal understanding or a series of email exchanges that do not constitute a contract, is in breach of section 8(2) (para 5.2).
Rule 6(1)(f) then requires appropriate provision in that contract for taking reasonable security safeguards. A contract that engages a Data Processor but is silent on security safeguards does not satisfy Rule 6(1)(f), even if it is otherwise a valid contract under the general law of contract (para 5.2).
On scope, the book resolves an interpretive point in favour of caution. Section 8(2) is limited by the phrase “for any activity related to offering of goods or services to Data Principals”, and an engagement for purely internal administrative processing, such as a payroll vendor for the Data Fiduciary’s own employees, would arguably fall within it through the employment-related legitimate use under section 7(i). The safer interpretation, and the one a compliance programme should adopt, is that every Data Processor engagement involving personal data is formalised through a valid contract with security provisions (para 5.2).
For the role question, one clause matters more than the rest: the processor’s duty to notify the Data Fiduciary without delay on becoming aware of a personal data breach, and to cooperate with its notification obligations under Rule 7 (paras 5.7.2 and 6.1.6). The seventy-two-hour Board reporting clock runs from the Data Fiduciary becoming aware, so a processor that notifies late makes the Data Fiduciary’s timeline unachievable, and section 8(1) leaves the Data Fiduciary accountable for the result.
The full clause set, the handling of sub-processors and what to check in a vendor’s standard terms are covered in the post on the data processing agreement under the DPDP Act.
Beyond the contract sits the oversight. Section 8(1) makes the Data Fiduciary responsible for processing undertaken on its behalf, so the accountability framework must extend to vendor oversight: a register of all Data Processors, monitoring of their compliance with the Rule 6(1)(f) security provisions, periodic assessments of their security posture, and confirmation that each can support the Data Fiduciary’s obligations on breach notification, erasure and Data Principal rights. The book’s warning is worth carrying into a procurement review: organisations that treat vendor management as a procurement function rather than a compliance function will find that the accountability gap between the contract and the practice is where enforcement risk accumulates (para 5.7.1).
Where the vendor register and the periodic assessment programme have to be built rather than described, including compliance SOPs and automation of compliances, Taxmann Advisory provides implementation support.
9. What obligations does a Data Processor actually bear?
Three, and they are frequently underestimated by the processor itself (para 2.4):
| Obligation | Source |
|---|---|
| Process personal data in accordance with the contract entered into with the Data Fiduciary | Section 8(2) |
| Implement reasonable security safeguards as provided in the contract | Rule 6(1)(f) |
| Erase personal data when directed to do so by the Data Fiduciary | Section 8(7)(b) |
Two further points attach to that list.
Rule 8(3) reaches processor-side processing directly. It requires the retention of personal data, associated traffic data and other logs of processing for a minimum period of one year from the date of such processing, in respect of any processing undertaken by the organisation or on its behalf by a Data Processor (Appendix A, section 7). The log retention obligation therefore has to be satisfied across the processor estate, not only in the Data Fiduciary’s own systems.
And the erasure obligation has to be evidenced on both sides. The book’s audit test for this area asks whether Data Processor erasure under section 8(7)(b) is evidenced by erasure instructions and confirmations, and lists the recurring finding plainly: Data Processor erasure not tracked, with no instructions issued and no confirmations received (para 8.10).
The misunderstanding the book records in its sector findings is the broadest version of this error: SaaS providers treating themselves as having no DPDP obligations because they are Data Processors, overlooking the security safeguard, contractual and erasure obligations that do apply to Data Processors (para 13.10).
10. What does an auditor test, and what goes wrong?
On role classification, Appendix A frames the requirement as a testable statement: the organisation shall identify and document its role under the Act, whether it operates as a Data Fiduciary determining the purpose and means of processing, a Data Processor processing on behalf of a Data Fiduciary, or both in different contexts (Appendix A, section 1).
On processor management, the book’s audit test asks whether every Data Processor engagement is formalised through a valid contract containing the security provisions required by Rule 6(1)(f) (para 5.8). The evidence to have ready: a Data Processor register listing all engaged processors with contract status, and sample Data Processing Agreements demonstrating the Rule 6(1)(f) security clauses (para 5.8).
The findings that recur across these two areas:
| Finding | Source |
|---|---|
| Data Processor engagements without written contracts, particularly with legacy vendors or informal arrangements | para 5.8 |
| Contracts that cover commercial terms but omit security safeguard provisions | para 5.8 |
| Processing activities without any documented lawful basis, arising most frequently with processing performed by third-party vendors under informal arrangements | para 3.8 |
| Data Processor erasure not tracked, with no instructions issued and no confirmations received | para 8.10 |
| Cross-border transfers occurring through cloud replication or content delivery network caching that are not documented in the transfer register | para 8.10 |
| SaaS providers treating themselves as having no DPDP obligations because they are Data Processors | para 13.10 |
The cloud replication finding is the one that defeats a register built from the contract rather than from the architecture. A transfer the organisation never negotiated, because it happens automatically inside a provider’s platform, is still a transfer, and it is invisible unless the data flow mapping is done against the systems rather than against the vendor list.
11. Quick answers
What is the difference between a Data Fiduciary and a Data Processor?
A Data Fiduciary, under section 2(i), is any person who alone or in conjunction with other persons determines the purpose and means of processing of personal data. A Data Processor, under section 2(k), is any person who processes personal data on behalf of a Data Fiduciary. The test is control: who decides why the data is processed and how, not who physically handles it. The Data Fiduciary bears the primary obligations under sections 4 to 10; the Data Processor’s obligations are principally contractual, security-related and erasure-related.
Can a company be both a Data Fiduciary and a Data Processor?
Yes, and commonly is. Taxmann’s Digital Personal Data Protection gives the example of an IT services company that is a Data Fiduciary for the employee personal data it processes for its own HR purposes and a Data Processor for the client customer data it processes under a service agreement. The classification must be determined at the level of each processing activity, not at the organisational level.
Is a cloud provider a Data Processor under the DPDP Act?
It depends on the activity. A cloud infrastructure provider offering storage and compute resources, where the customer determines what data is stored and how it is processed, is ordinarily a Data Processor. A cloud-based CRM platform that determines the structure of data fields, the analytics algorithms applied to customer data and the retention policies may cross into Data Fiduciary territory for some processing activities, because those are decisions about purpose and means.
Does the DPDP Act recognise joint controllers?
Not by that name. The DPDP Act does not use the term “joint controller” or “co-fiduciary”, but the phrase “alone or in conjunction with other persons” in section 2(i) creates the same conceptual space. Where two or more entities jointly determine the purpose and means of processing, each is a Data Fiduciary for that processing and bears the full set of obligations. How liability is allocated between them is not expressly addressed in the Act.
Can a contract transfer a Data Fiduciary’s obligations to a vendor?
No. Section 8(1) provides that the Data Fiduciary is responsible for complying with the Act and the Rules in respect of any processing undertaken by it or on its behalf by a Data Processor, irrespective of any agreement to the contrary. As Taxmann’s Digital Personal Data Protection puts it, the Data Fiduciary may delegate the processing but cannot delegate the accountability.
What should a role classification register record?
Five fields for each processing activity: a description of the activity and the categories of personal data involved; the identity of the entity or entities that determine the purpose of processing; the identity of the entity or entities that determine the means; the classification as Data Fiduciary, Data Processor or joint Data Fiduciary; and the date and authorship of the assessment. The register should be reviewed annually at a minimum, and immediately on any material change in processing activities.
Do Data Processors have obligations under the DPDP Act?
Yes. A Data Processor must process personal data in accordance with the contract entered into with the Data Fiduciary under section 8(2), implement reasonable security safeguards as provided in the contract under Rule 6(1)(f), and erase personal data when directed to do so by the Data Fiduciary under section 8(7)(b). Rule 8(3) also reaches processing undertaken on a Data Fiduciary’s behalf, requiring one-year retention of personal data, associated traffic data and processing logs. Taxmann’s Digital Personal Data Protection records, among its sector findings, SaaS providers wrongly assuming they have no DPDP obligations at all because they are processors.
This article draws on Chapter 2 of Taxmann’s Digital Personal Data Protection – An Essential Guide to India’s DPDP Law by CA Narasimhan Elangovan (2026 Edition), which defines the Data Fiduciary, Data Processor and Data Principal, examines the control test and joint determination, and sets out the role classification exercise and the five fields a role determination must document. It also draws on Chapter 5 on the engagement of Data Processors under section 8(2) and the minimum clause checklist in Figure 5.1, Chapter 6 on the contractual security provisions required by Rule 6(1)(f), Chapter 8 on Data Processor erasure, Chapter 12 on where role classification sits in the first ninety days, and Appendix A on the evidence each determination should leave behind. Law stated as updated till 15 September 2026.
The DPDP Act, 2023, the DPDP Rules, 2025 and the notifications of 13 November 2025 are available on Taxmann.com | Research under Indian Acts and Rules.
Related reading: scope and key definitions under the DPDP Act · obligations of Data Fiduciaries · cross-border data transfers · the DPDP Act compliance checklist for businesses.
More in this series: the Data Protection Board of India · how a DPDP compliance audit is conducted · Significant Data Fiduciary obligations · Consent Managers under the DPDP framework · personal data breach intimation under Rule 7 · handling Data Principal rights requests · children’s data and verifiable parental consent · the SPDI Rules and the DPDP Act · the Record of Processing Activities (ROPA) · the incident response plan · data retention and erasure · the privacy notice under Rule 3 · the data processing agreement · data localisation and cross-border transfers · employee data and workplace monitoring · exemptions under the DPDP Act · legitimate uses under section 7.
Disclaimer: The content/information published on the website is only for general information of the user and shall not be construed as legal advice. While the Taxmann has exercised reasonable efforts to ensure the veracity of information/content published, Taxmann shall be under no liability in any manner whatsoever for incorrect information, if any.

Taxmann Publications has a dedicated in-house Research & Editorial Team. This team consists of a team of Chartered Accountants, Company Secretaries, and Lawyers. This team works under the guidance and supervision of editor-in-chief Mr Rakesh Bhargava.
The Research and Editorial Team is responsible for developing reliable and accurate content for the readers. The team follows the six-sigma approach to achieve the benchmark of zero error in its publications and research platforms. The team ensures that the following publication guidelines are thoroughly followed while developing the content:
- The statutory material is obtained only from the authorized and reliable sources
- All the latest developments in the judicial and legislative fields are covered
- Prepare the analytical write-ups on current, controversial, and important issues to help the readers to understand the concept and its implications
- Every content published by Taxmann is complete, accurate and lucid
- All evidence-based statements are supported with proper reference to Section, Circular No., Notification No. or citations
- The golden rules of grammar, style and consistency are thoroughly followed
- Font and size that’s easy to read and remain consistent across all imprint and digital publications are applied


CA | CS | CMA