Naavi.org

The DGPSI-AI is a framework conceived for use by deployers of AI who are “Data Fiduciaries” under DPDPA 2023.

An interesting set of observations have been received recently from a professional regarding the framework. We welcome the comments as an opportunity to improve the framework over a period of time. In the meantime, let us have an academic debate to understand the concerns expressed and respond.

The observer  had made the following four observations as concerns related to the DGPSI-AI framework.

1. AI’s definition is strange and too broad. Lots of ordinary software has adaptive behavior (rules engines, auto-tuning systems, recommender heuristics, control systems). If you stretch “modify its own behavior,” you’ll start classifying non-AI automation as AI. Plus, within AI spectrum, only ML models may have self learning capabilities. Linear and statistical and decision tree models do not.

2. “AI risk = data protection risk = fiduciary risk”. That is legally and conceptually incorrect. DPDP Act governs personal data processing, not AI behavior as such. Many AI risks cited (hallucination, deception, emergent behavior, hypnosis theory) are safety / reliability / ethics risks, not privacy risks.

3. Unknown risk = significant risk” is a logical fallacy. Unknown ≠ high. Unknown risk can be negligible, bounded or mitigated through controls. Risk management is about estimating and bounding uncertainty,

4.Explainability treated as a legal obligation, not a contextual requirement. This is overstated. DPDP requires notice, not model explainability.

I would like to provide my personal response to these observations, as follows:

1. AI Definition

DGPSI has recommended adoption of a definition for AI which reflects an ability for automated change of the execution code based on the end results of the software without an intervention of a human for creating a modified version.

The “Rules Engine”, “auto tuning systems” or such other systems that may be part of the ordinary software are characteristic by the existence of a code for a given context and situation. If the decision rule fails, the software may either crash or use a default behaviour.  The outcome is therefore not driven by a self learning of the software. It is pre-programmed by a human being. Such software may have higher degree of automation than most software but need not be considered as AI in the strict sense.

Therefore, iff there is any AI model where the output is pre-determined, it can be excluded from the definition of AI by a DGPSI-AI auditor with suitable documentation.

Where the model self corrects and over a period of time transforms itself like metamorphosis into a new state, without a human intervention, the risk could be that further outputs may start exhibiting more and more hallucinations or unpredictable outcomes. The output data which may become input data for further use may get so poisoned that the difference between reality and artificial creation may vanish. Hence such behaviour is classified as AI.

In actual practice, we tend to use the term “AI” loosely to refer to any autonomous software with a higher degree of autonomy. We can exclude them from this definition. The model implementation specification MIS-AI-1 in the framework states as follows:

“The deployer of an AI software in the capacity of a Data Fiduciary shall document a Risk Assessment of the Software obtaining a confirmation from the vendor that the software can be classified as ‘AI’ based on whether the software leverages autonomous learning algorithms or probabilistic models to adapt its behaviour and generate outputs not fully predetermined by explicit code. This shall be treated as DPIA for the AI process”

This implementation specification which require documentation for the purpose of compliance, may perhaps address the concern expressed.

2. AI Risk and Privacy Risk

The framework DGPSI-AI is presented in the specific context of a responsibility of a “Data Fiduciary” processing “Personal Data”.

Since non compliance of DPDPA leads to a financial risk of Rs 250 crores+, it is considered prudent for the data fiduciary to consider AI behavioural risks as risks that can lead to non compliance.

In the context of our usage, hallucination, rogue behaviour, etc which are termed “Safety” or “Ethics” related issues in AI are applied for recognizing “Unauthorized processing of personal data” and hence become risks that may result in hefty fines. We cannot justify with the Data protection board that the error happened because I was using AI and hence I must be excused.

Hence AI risks become Privacy Risks or DPDPA Non Compliance Risks.

3. Unknown Risk:

The behaviour of AI is by design  meant  to be  creative  and therefore is unpredictable. All Risks associated with the algorithm is not known even to the developer himself. They are definitely to be classified as “Unknown Risks” by the deployer.

We accept that Unknown Risk can be negligible. But we come to know of it only after the risk becomes known. A Fiduciary cannot assume that the risk when determined will be negligible. If he has to determine if he is a “Significant Data Fiduciary” or not, he should be able to justify that the risk is negligible ab-initio. This is provided for in the framework by MIS-AI-3 which suggests,

“Where the data fiduciary in its prudent evaluation considers that the sensitivity of the “Unknown Risk” in the given process is not likely to cause significant harm to the data principals, it shall create a “AI-Deviation Justification Document” and opt not to implement the “Significant Data Fiduciary” obligations solely as a reason of using AI in the process. “

This provides a possibility of “Absorbing” the “Unknown Risk” irrespective of its significance including ignoring the need to classify the deployer as a “Significant Data Fiduciary”.

Hence there is an in-built flexibility that addresses the concern.

4.Explainability

The term “Explainability” may be used by the AI industry in a particular manner. DGPSI-AI tries to use the term also to the legal liability of a data fiduciary to give a clear, transparent privacy notice.

A “Notice” from a “Fiduciary” requires to be clear, Understandable and transparent by the data principal and hence there is a duty for the Data Fiduciary to understand the AI algorithm himself.

It may not be necessary to share the Explainability document of the AI developer with the data principal in the privacy notice. But the Data Fiduciary should have a reasonable assurance that the algorithm does not cause any harm to the data principal and its decisions are reasonably understood by the Data Fiduciary.

Towards this objective, MIS-AI-6 states:

“The deployer shall collect an authenticated “Explainability” document from the developer as part of the licensing contract indicating the manner in which the AI functions in the processing of personal data and the likely harm it may cause to the data principals.”

I suppose this reasonably answers the concerns expressed. Further debate is welcome.

Naavi

Mr Vinod  Sreedharan is a AI expert with a creative bent of mind. He has applied his creative thoughts to give a visual imagery touch to “Taming of DPDPA and AI with DGPSI-AI”

The complete document above in PDF format is available here 

The Meity has released some corrections in DPDPA Rules through a Gazette notification

It essentially consists of some typo corrections.  as follows

Necessary corrections are being made at dpdpa.in and the rules posted at the dpdpa rules 

Naavi

The Government of Karnataka has recently passed a bill titled “The Karnataka Hate Speech and hate Crimes (Prevention) Bill 2025 (LA Bill No 79 of2025) . It is currently  pending Governor’s assent.

The first thing we note in this bill is that it covers both “Speech” and “Crime” and includes “Electronic Form”.

Regulating the speech  is subject to restrictions in the constitution. A law to directly curtail speech is ultra vires the Constitution Article 19.

Legislation on “Electronic  Documents” is under Information Technology Act  2000 (ITA 2000) and the powers of the State Government to legislate for the use of Electronic document is restricted to Section 90 of the ITA 2000. It does not extend to creating new Cyber Crimes under the power of the State.

Section 6 of the Act provides power to block the hate crime materials and if it is in electronic form, directly conflicts with Section 69 and 69A of the Information Technology Act 2000 (ITA 2000) and rules made there under which places such powers  with several restrictions.

For ease of understanding, we quote Section 90 of ITA 2000 here

Section 90: Power of State Government to make rules

(1) The State Government may, by notification in the Official Gazette, make rules to carry out the provisions of this Act.

(2) In particular, and without prejudice to the generality of the foregoing power, such rules may provide for all or any of the following matters, namely –

(a) the electronic form in which filing, issue, grant receipt or payment shall be effected under sub-section (1) of section 6;
(b) for matters specified in sub-section (2) of section 6;

(3) Every rule made by the State Government under this section shall be laid, as soon as may be after it is made, before each House of the State Legislature where it consists of two Houses, or where such Legislature consists of one House, before that House.

It is clear that the powers of the State Government to legislate on the use of Electronic Documents for Governance is restricted and does not extend  to defining new Cyber Crimes and prescribing punishments.

This is considered void ab-initio.

Hence  including “Electronic” form of communication under the Bill renders the Bill illegal and liable to be struck down.

The world of Electronic Documents constitutes what is generally described as the “Cyber Space” which is recognized as an independent area of activity. Countries such as USA recognize “Cyber” as a separate command for their defence forces exactly for the reason that it represents an extension of the geographical boundaries of the sovereign country similar to the sea and air space.

Hence law of cyber space does not fall under the concurrent list and is the sole domain of the Central Government to legislate. This legislation and all legislations so far in multiple States are therefore unconstitutional and needs to be stuck down. 

A suitable petition constitutional bench in the Supreme Court needs to be considered to decide about the status of “Cyber Space” and the rights of a sovereign entity to draw Cyber boundaries and legislate for crimes there in.

Currently ITA 2000 already has established such boundaries and conditions when the extra territorial jurisdictions can be  extended to foreign territories. Such powers  under Section 75 cannot be left to be exercised by State Government and hence the Karnataka Hate Act cannot include “Electronic Documents” as instruments by which offences under the Act may be committed.

Hence there is a need to omit “Electronic Documents” from the provisions of the Karnataka Hate Bill under Section 2(i) and 2(iv).

Hence the part of the Bill that curtails speech  particularly in the electronic form is considered unconstitutional and ultravires the Indian Constitution as well as ITA 2000.

Additionally,  the punishment envisaged under the act even for the first time offender is “Imprisonment of not less than one year which may extend to seven years” and for subsequent offences can extend upto 10 years.  The offence is “Cognizable”, “Non Bailable”  and “triable by the Judicial magistrate First Class”.

Hence the offence is graded as “Heinous” and can be grossly abused by the Police. It can therefore have a “Chilling Effect” as the Supreme Court defines under the “Shreya Singhal Case”.

There is therefore an urgent need for the Bill to be withdrawn by the Assembly,  failing which to be rejected by the Governor, failing which to be struck down by the appropriate Courts.

This issue being a serious Constitutional matter,  has to be taken up by some public spirited law firm and fought in Karnataka High Court and the Supreme Court.

I hope  such people take note.

Comments are welcome.

Copy of the Bill

Refer:

https://www.youtube.com/watch?v=3YmexzlaPko

Naavi

The launch of DGPSI-GDPR is not  just another event. It is a symbol of Indian Data Protection eco system coming of age not only to be compliant with  DPDPA 2023 but also provide the guidance to the compliance of other data protection laws such as GDPR. The work has started and with the cooperation of the community we will have a framework that is acceptable as a good guidance to all companies firstly in India.

DGPSI (Digital Governance and Protection Standard of India) was developed as a guidance framework for compliance of DPDPA. It is a useful framework today for implementation of the DPDPA 2023 in an organization as well as for audit and assessment.

In India we also have many organizations who process data from outside India and most of them so far treated GDPR as the standard for Data Privacy Compliance. With the coming of ISO 27701:2025, the GDPR Compliance through ISO 27701:2025 as an independent certifiable framework also received a boost.

In this context, most organizations in India are confronted with the need to look at two compliance drives  one for DPDPA and another for GDPR.

While some would like to adapt GDPR compliance to DPDPA compliance and use ISO 27701:2025 (modified for India), an alternative is to use DGPSI and adapt it to GDPR compliance.

To facilitate this use of a Made in India framework for compliance of GDPR, DGPSI has now been extended with a DGPSI-GDPR version. This uses the 50 Model implementation specifications of DGPSI with subtle changes to be capable of meeting the GDPR requirements.

This is a a game changer in the domain of Data Protection Compliance in India and a transition point where DGPSI becomes the source framework from which compliance of Data Protection laws of other countries can be carved out.

Currently, FDPPI is working on a draft version of the DGPSI-GDPR version and the Certified Data Auditors of FDPPI will be trained to use the version for GDPR compliance as may be required.

Under DPDPA, data processing activities where process foreigner’s data under a contract are exempted from DPDPA. Such activities involving  EU data are now capable of being implemented and audited using DGPSI-GDPR. It is one of the requirements of DGPSI that personal data is classified on the basis of applicable jurisdiction and hence even where the data is currently mixed up, they need to  be segregated and a virtual GDPR processing division has to be created. Such  virtual division can now use DGPSI-GDPR as the framework for compliance.

Can an India framework take on the compliance of Global Data Protection Compliance? …will be a question in the minds of many data protection professionals.

Let us make it happen. FDPPI invites all data protection professionals in India to put in their efforts to develop the DGPSI family of frameworks to expand and provide guidance to the compliance of GDPR as well as other data protection laws in due course.

Naavi

FDPPI’s Certification program for developing DPOs and Data Auditors in India offers an online  examination for professionals to validate their knowledge and skills to be a good DPO in the Indian scenario.

While trainings are conducted by the  training partners of FDPPI from time to time (Eg: Virtual Program on December 20-21 conducted by Cyber Law College), the coveted certification of C.DPO.DA. is available  for anybody who registers for the online examination and pass through  the  cut-off marks required.

Whether you are a CIPP certified or DSCI certified or ISO certified or PECB certified or EXIN Certified or ISACA Certified you can appear for this online examination by making payment  of the prescribed fees and take the examination.

As a special year end offer, the examination for which the fee is normally Rs 25000/- is being offered at Rs 10000/- till 31st December 2025.

The material for the exam is available in the following three books

The Certification training conducted by Cyber Law College consists of 12 hours of online discussions that cover

• Legal nuances of DPDPA and the DPDPA  Rules
• DPDPA Risk Assessment and Case Study
• The Roles of DPO and Data Auditor in the DPDPA era
• Classification of DPDPA protected Data (DPD) and ROPA as a strategic tool of Compliance
• DPDPA Compliance by Default:  Technical challenges and Designing Controls
• Use of DGPSI as a Compliance Management framework
  • DGPSI Full, DGPSI-Lite and DGPSI-AI
  • DGPSI-GDPR,DGPSI HR, DGPSI Data Processor
• Comparison of DGPSI with ISO 27701
• Discussions

Besides training themselves to be DPOs in an organization, some of our trainees may emerge as independent trainers in different parts of the country under a franchise scheme of FDPPI.

FDPPI is also introducing an upgrade over C.DPO.DA. for those who want to be “Independent Data Auditors” which is a position that is likely to open up in 2027 after the Act becomes fully effective. A separate upgradation training for CIDA (Certified Independent Auditor) is being planned to be conducted in 2026 for this purpose.

For the time being it is an opportunity for interested professionals to take the C.DPO.DA. examination at the special year end price of Rs 12000/-. (Inclusive of GST)  You can register directly on the CDPODA page of fdppi.in and making a payment of Rs 12000/-. (Indicate in the description that the registration is for examination only)

Naavi