Lokesh Bhardwaj Law Offices ← Articles
Lokesh Bhardwaj Law Offices · Advocates and Legal Consultants, New Delhi · lokeshbhardwajlawoffices.com
Technology Law · Article

Data without borders, liability without limits

The modern enterprise's guide to data custody: the regimes of India, the EU and the US, the cost of disregarding them, and whether data at scale can still be protected at all.

13 June 202622 minute read
Last reviewed: 17/10/2026
ASSET owned and exploited THE BORDER TRUST held in custody the data is held, not owned; the obligation crosses the border with the person
Personal data is held, not owned. The obligation attaches to the person and crosses the border with the data.

Data has become the substance of modern enterprise, not a by-product of it. The personal data a business holds is never quite its own, and the law that protects it now follows that data across every border it crosses. What follows is a study of the regimes governing India, the European Union and the United States, the cost of disregarding them, and the harder question the age of artificial intelligence has forced upon us: whether data at scale can still be protected at all.

There is a quiet contradiction at the centre of the data economy, and most of the disputes, penalties and compliance failures of the last decade can be traced to it. A business treats the data it gathers as an asset: something it owns, banks, leverages and, when convenient, sells. The law treats that same data, where it describes a living person, as something the business merely holds in trust, encumbered by the continuing rights of the individual it concerns. The enterprise is not the owner. It is the custodian. That distance between possession and ownership, between the instinct to exploit and the duty to protect, is what may be called the custodial paradox, and it widens with every year that data grows more valuable and more mobile.

The paradox would be a matter of philosophy were it not enforced with fines now measured in the billions, with orders that can halt a company's operations overnight, and with a body of cross-border law that asserts jurisdiction over data the moment it leaves the territory in which it was collected. For any enterprise that handles personal information at scale, and for the advocate who must advise it, understanding this is no longer optional. It is the difference between a defensible data strategy and a contingent liability sitting quietly on the balance sheet.

I. The substance, not the residue, of the business

It has become a cliche to say that data is the new oil, and like most cliches it is both true and misleading. Oil is consumed when used; data is not. Data can be copied at no marginal cost, combined with other data to yield insight neither set held alone, and inferred from behaviour the individual never consciously disclosed. A retailer once kept records about its trade. The modern platform is its data: its valuation, its competitive moat, its capacity to predict and to personalise, all reduce to the corpus it has assembled and the models it has trained on that corpus.

This is why the legal character of data has moved from the periphery of commercial practice to its core. When data was a record of the business, data protection was a back-office hygiene matter. When data becomes the business itself, data protection becomes a question of title, of saleability, of cross-border mobility and of strategic risk, which is to say a question for the boardroom and for counsel rather than for the IT department alone. The first error most enterprises make is to treat it as the latter when it has long since become the former.

II. Asset or trust? The classification that decides everything

Before one asks how data may be protected, sold, or transferred, one must ask what kind of data it is and what kind of business is holding it, because the answers turn entirely on classification. The law does not regulate data in the abstract; it regulates particular categories of data held for particular purposes by particular kinds of actor.

The first cut is between personal and non-personal data. Personal data, broadly, is any information relating to an identified or identifiable individual. Within it sits a more sensitive stratum, treated with heightened care in nearly every modern regime: data revealing health, biometrics, genetics, religious or political belief, sexual orientation and the like, what the General Data Protection Regulation calls special categories under Article 9, and what India's earlier rules under the Information Technology Act treated as sensitive personal data. Outside personal data lies non-personal, aggregated and genuinely anonymised data, which carries far lighter obligations. The second cut is the corporate one familiar from information-security practice: public, internal, confidential and restricted. The third is sectoral: financial data, health records, telecom subscriber data and payment data each attract their own specialised regimes layered on top of the general law.

Classification is not a clerical exercise. It determines your lawful basis for holding the data, your retention period, your security obligations, whether you may send it abroad, and, critically, whether you may sell it at all. Which returns us to the asset question, and to the most expensive misconception in the field.

The data you hold is valuable precisely because it describes people. For that same reason it is not freely yours to sell.

Personal data is not property you own and may alienate like a chattel. Under India's Digital Personal Data Protection Act, 2023, the entity that determines the purpose and means of processing is a Data Fiduciary, and the word is chosen with care: a fiduciary holds for the benefit of another and is accountable to that other. Under the GDPR the equivalent actor is a controller, bound to a lawful basis for every use and to the data subject's rights of access, correction, erasure and objection. Under California's regime the position is sharper still: the statute regulates the sale and the sharing of personal information as distinct, named activities, and gives the consumer a standing right to opt out of both. Honouring that opt-out is not a courtesy; its breach is enforced, and a recent opt-out failure produced a settlement of USD 1.35 million, a figure itself surpassed in February 2026 by a settlement exceeding USD 2.75 million.

So the honest answer to how data is to be sold is that personal data, as such, very often may not be, not without a lawful basis, transparency and, in many corridors, the individual's freely given consent and a continuing right to withdraw it. What may be commercialised cleanly is non-personal, aggregated or truly anonymised data. The qualifier truly is doing heavy work, because anonymisation is far more fragile than the marketing literature suggests: re-identification through the combination of supposedly innocuous fields is a well-documented attack, and a dataset that is anonymous today may become identifying tomorrow when joined with a set that did not exist when it was released. The enterprise that sells anonymised data and is later shown to have sold re-identifiable personal data has not sold an asset; it has manufactured a liability and distributed it to every buyer.

What kind of business you are in therefore decides the entire analysis. For a data broker, personal data is the product, and the regulatory burden is correspondingly central and existential. For a hospital, a bank or an insurer, it is the lifeblood of a regulated activity, governed by sectoral rules stricter than the general law. For an ordinary enterprise, it is a by-product of operations that must be guarded but should rarely be sold. The same byte attracts radically different duties depending on whose hands hold it and why.

III. The cartography of a borderless thing

Data does not respect frontiers. A keystroke in Delhi may rest on a server in Virginia, be processed by a model in Ireland and be backed up in Singapore, all within a second and without the individual's awareness. Law, by contrast, is territorial: a sovereign legislates for its own territory and, increasingly, for its own residents wherever their data may travel. The collision of a borderless medium with a territorial law is the cross-border data problem in a sentence, and it is the reason the single most consequential penalties in this field have concerned not how data was secured but where it was sent.

Modern regimes resolve the collision by attaching the protection to the person and following the data outward. The GDPR applies extraterritorially to any organisation, anywhere, that processes the personal data of people in the European Union in connection with offering them goods or services or monitoring their behaviour. India's DPDP Act reaches processing outside India where it is in connection with offering goods or services to people within India. The practical consequence is that a firm in Gurugram or in Texas may be bound by European law, and a firm in Berlin may be bound by Indian law, with no physical presence in the other's territory at all. There is no exemption for company size or turnover. The reach is the point.

IV. Three architectures of protection

The three regimes that matter most to an Indian enterprise trading globally are the European, the American and the Indian, and they are built on three different philosophies. It is worth setting them side by side.

Europe: rights as the default

The European Union treats data protection as a fundamental right, given concrete form by the GDPR (Regulation (EU) 2016/679) since 2018. Its architecture is principle-based and rights-centred: a lawful basis is required for every processing activity; data must be minimised, kept only as long as necessary and secured; and the individual holds enforceable rights against the controller. Transfers out of the European Economic Area are tightly governed by Chapter V, of which more below. Layered onto this, the European Union has now added the world's first comprehensive statute for artificial intelligence, the AI Act (Regulation (EU) 2024/1689), which entered into force on 01/08/2024 and applies in stages. Its prohibitions and literacy obligations have applied since 02/02/2025; obligations for general-purpose AI models since 02/08/2025; and the substantial body of obligations for high-risk systems and for transparency from 02/08/2026, a date now weeks away as this is written. The AI Act's maximum penalty, up to 35 million euros or 7 per cent of worldwide annual turnover, exceeds even the GDPR's ceiling, and the two regimes operate in tandem: data-protection authorities retain jurisdiction over the personal data an AI system consumes, while market-surveillance authorities police the AI Act itself. An enterprise deploying AI in Europe therefore answers to both at once.

The United States: a sectoral and state-by-state patchwork

The United States has, to this day, no single comprehensive federal privacy statute on the European or Indian model. The most recent attempt, the American Privacy Rights Act introduced in 2024, advanced in committee and then stalled; as matters stand no federal omnibus law has passed. Federal protection is instead sectoral: health information under HIPAA, financial information under the Gramm-Leach-Bliley Act, children under 13 under COPPA, education records under FERPA, and a general prohibition on unfair or deceptive practices enforced by the Federal Trade Commission. Into the federal vacuum the states have moved with force. By 2026 roughly 20 states have enacted comprehensive consumer-privacy laws, beginning with California's CCPA and its successor the CPRA, with Indiana, Kentucky and Rhode Island among the latest to take effect; 12 states now legally require honouring a universal browser opt-out signal, and nearly all require opt-in consent for sensitive data. The result is a compliance map that changes by the visitor's location, and enforcement that is no longer theoretical: reported state penalties reached an estimated USD 1.4 billion in 2025 alone, with one technology company settling with the State of Texas for a sum exceeding USD 1 billion. The American position is best described not as light-touch but as fragmented and intensifying.

India: a rights regime with a sovereign hand

India arrived at a comprehensive statute through the front door of the Constitution. The right to privacy was affirmed as a fundamental right in Justice K. S. Puttaswamy v. Union of India, (2017) 10 SCC 1, and the Digital Personal Data Protection Act, 2023 is its statutory expression. After a long gestation, the operative subordinate legislation, the Digital Personal Data Protection Rules, 2025, was notified on 13/11/2025 (G.S.R. 846(E)) under Section 40 of the Act, and the framework now comes into effect in phases: foundational and Board provisions at once; the consent-manager regime around 13/11/2026; and the substantive obligations, including the cross-border provisions of Section 16, roughly 18 months from notification, around 13/05/2027. Enforcement sits with a fully digital Data Protection Board of India of four members, with complaints filed and tracked online, and, in a detail of particular interest to practitioners, appeals from the Board lie to the Telecom Disputes Settlement and Appellate Tribunal. The penalty schedule is severe: up to Rs. 250 crore for a failure to take reasonable security safeguards that results in a breach, up to Rs. 200 crore for a failure to notify a breach or for lapses in protecting children's data, and up to Rs. 50 crore for other contraventions. India has built a genuine rights regime; it has also, as we shall see, reserved a wide hand for the State within it.

And the rest of the world

These three are not the whole map. More than 140 jurisdictions now have data-protection or privacy legislation on their statute books, and the gravitational pull of the European model is visible across most of them. Brazil's LGPD, in force since 2020 and policed by its national authority, is closely modelled on the GDPR. China has taken a markedly more state-centred path, combining its Personal Information Protection Law of 2021 with a Data Security Law and a Cybersecurity Law, classifying data into core, important and general tiers, requiring localisation for operators of critical information infrastructure, and conditioning many outbound transfers on a security assessment by the cyberspace regulator. The lesson for the Indian exporter of data is plain: there is no single global standard to comply with, only a thickening lattice of overlapping regimes, and the corridor between any two countries must be assessed on its own terms.

V. Moving data lawfully across the line

If protection follows the data outward, then the central operational question for any cross-border business is: by what mechanism may I lawfully send personal data from country A to country B? Here the regimes diverge sharply, and the divergence is itself a compliance trap.

Europe operates a whitelist logic. A transfer outside the European Economic Area is lawful only if the destination has been found by the European Commission to offer adequate protection, or if the exporter puts in place an approved safeguard: Standard Contractual Clauses, Binding Corporate Rules for intra-group transfers, or, narrowly, one of the derogations for specific situations. The adequacy route is where the long transatlantic drama has played out. The European Court invalidated the first United States arrangement, Safe Harbour, in the Schrems I decision of 2015 (Case C-362/14), and the second, Privacy Shield, in Schrems II in 2020 (Case C-311/18), each time on the ground that United States surveillance gave European data insufficient protection and redress. The third arrangement, the EU-US Data Privacy Framework, was adopted by adequacy decision in July 2023 and survived its first courtroom test when the General Court dismissed the challenge in Latombe v. Commission (Case T-553/23) on 03/09/2025. That is not the end of the matter: the applicant has appealed to the Court of Justice, and the saga that twice invalidated a transfer framework remains, formally, unfinished. Any business relying on the Framework should treat it as valid today and contingent tomorrow, and keep Standard Contractual Clauses on the shelf as a fallback.

India has chosen the opposite default: a negative list. Under Section 16 of the DPDP Act a Data Fiduciary may transfer personal data to any country or territory, freely, unless and until the Central Government, after assessment, restricts transfers to a named country by notification. As at the time of writing no such restricted-country list has been published. Rule 15 of the 2025 Rules adds that the fiduciary must comply with any conditions the Government prescribes, and is broadly worded enough to restrict the act of making personal data available to a foreign State or to entities under its control. The negative-list model is more permissive than Europe's on its face, but it carries two stings in the tail. First, it is executive-led and dynamic, so a corridor that is open today may be closed by notification with little warning; the prudent enterprise builds in the capacity to reroute. Second, and more immediately, the DPDP framework does not displace sectoral localisation. The Reserve Bank of India's 2018 directive requires that payment-system data be stored only on servers within India, and comparable expectations exist in the securities and insurance sectors. For a regulated entity, the sectoral rule is the binding constraint, and it is frequently stricter than the general law.

The single most expensive penalties in data protection have concerned not how data was secured, but where it was sent.

VI. The technology-transfer overlay

Cross-border data flows do not travel alone. They move alongside, and often inside, the transfer of technology, and a second and less familiar body of law has begun to merge with data-protection law: export control and national security. The convergence is now explicit. In the United States, Executive Order 14117 directs the restriction of bulk transfers of sensitive personal data, and certain government-related data, to designated countries of concern, treating mass data flows not as a privacy matter at all but as a strategic and national-security one, and the implementing programme has moved into force. The logic is the export-control logic: some categories of data, like some categories of technology, are too sensitive in aggregate to leave the country freely.

The same logic has long governed technology itself. Under the United States export regime, the disclosure of controlled technology or source code to a foreign national, even one standing in the same office, can constitute a deemed export requiring authorisation; encryption software has its own controls; and some States condition market access on source-code review or escrow, a demand that sits uneasily with the protection of trade secrets. India maintains its own controls on dual-use technology through the SCOMET list. And data-localisation requirements, wherever they appear, function in practice as non-tariff barriers, compelling the local establishment of infrastructure and the local processing of data as the price of doing business. For the enterprise transferring both a service and the data that powers it, the upshot is that two regimes now apply at once: the privacy regime asking whether the individual is protected, and the strategic regime asking whether the nation is. A transfer can be impeccable under the first and unlawful under the second.

VII. The price of getting it wrong

The repercussions of non-compliance are no longer reputational alone, and they are no longer modest. They fall in several layers, and the headline fines, large as they are, are often the least of them.

Begin with the fines, because they set the scale. Cumulative penalties under the GDPR passed 7.1 billion euros across more than 2,800 enforcement actions in the period since 2018, with roughly 1.2 billion euros imposed in 2025 alone and breach notifications running at record levels. The single largest, 1.2 billion euros against Meta in 2023, was imposed not for a leak but for the unlawful transfer of European users' data to the United States while relying on Standard Contractual Clauses that were held inadequate after Schrems II. Cross-border enforcement is not confined to the transatlantic corridor: in 2025 a fine of 530 million euros was imposed for the unlawful transfer of European user data to China. Under the AI Act the ceiling rises to 35 million euros or 7 per cent of worldwide turnover. Under the DPDP Act it is up to Rs. 250 crore per default. In the United States, state penalties reached an estimated USD 1.4 billion in 2025. These are not outliers; they are the established range.

Yet the fine is frequently survivable where the corrective order is not. When Meta was fined, it was also ordered to suspend the offending transfers and to bring its processing into compliance, which for a business built on transatlantic data flows is an operational threat of a wholly different order. A transfer ban does not dent the balance sheet; it can stop the business. Beyond the regulator lie further layers: criminal exposure for certain offences and, in some regimes, personal liability for directors and officers; contractual consequences, including indemnity claims, breach of data-processing agreements, loss of certifications and the termination rights of enterprise customers whose own compliance depends on yours; the collapse of customer trust, which compounds; and the simple, brutal cost of incident response, notification and remediation. The lawful basis for treating data protection as a board-level concern is not exhortation. It is this list.

VIII. The machine that eats data

All of the foregoing is complicated, perhaps transformed, by artificial intelligence, because AI does not merely use data; it consumes it at industrial scale and in ways that strain every protective concept the law has built. The brief I have set myself ends with the hardest question this field now poses, so let me put it plainly: in a world where models are trained on vast corpora, much of it personal data harvested from the open web, can data still be protected at all? The honest answer is: partially, imperfectly, and only if one abandons the fantasy of perfect protection for the discipline of governed risk.

Consider what AI does to the protective assumptions. We assumed anonymisation was a wall; large models have shown it to be a fence, susceptible to re-identification when enough signals are combined. We assumed data used for one purpose stayed within that purpose; a model trained on data can memorise and, under the right prompt, regurgitate fragments of its training set verbatim, and can be probed by membership-inference and model-inversion attacks to reveal whether and what a particular individual's data contributed. The very capabilities that make these systems valuable, their capacity to generalise from data, are the capabilities that make the data within them difficult to contain. This is why Europe has not relied on the GDPR alone but added the AI Act on top, and why the two now impose compound and overlapping duties on anyone deploying AI on personal data.

Can technology answer the problem technology has created? In part. A family of privacy-enhancing techniques exists and is maturing. Federated learning sends the model to the data rather than the data to the model, so that raw records need not be centralised. Differential privacy injects calibrated mathematical noise so that the presence or absence of any one individual cannot be detected in the output. Secure multi-party computation and confidential computing, the latter using trusted hardware enclaves, allow joint computation without exposing the inputs. Homomorphic encryption, the most tantalising of all, permits computation directly on encrypted data, though it remains slow and costly for general use. Synthetic data and tokenisation reduce the exposure of the real thing. None of these is a silver bullet; each trades away utility, speed or cost, and each must be chosen for the specific risk it addresses. They are the instruments of governed risk, not of perfect safety, because perfect safety at this scale does not exist.

IX. The decryptable promise and the State's hand

Which brings us to encryption, and to a truth the marketing of security obscures. We speak of encryption as though it were an absolute. It is not. Encryption protects data convincingly in two states, at rest and in transit, but data must be decrypted to be processed, and in that moment of use it is plaintext. Keys can be lost, stolen or compelled. And a longer shadow has fallen across the field: the prospect that data harvested and stored today, while strongly encrypted, may be decrypted in future by quantum computation, the strategy adversaries call harvest now, decrypt later. It is no accident that new post-quantum cryptographic standards were finalised in 2024 and that migration to them has become a board-level concern. The promise of encryption is real but conditional, and the condition is time.

The sharpest illustration of the limits of encryption is not technical but political, and it concerns the access the State reserves to itself. In early 2025 the United Kingdom, acting under its Investigatory Powers Act of 2016, secretly served Apple with a technical capability notice demanding the ability to access data protected by the company's end-to-end-encrypted cloud service, with a statutory gag forbidding disclosure of the demand. Apple's response was instructive: rather than build the access, which cannot be built for one government without weakening the encryption for every user, it withdrew the feature from the United Kingdom on 21/02/2025 and challenged the order before the Investigatory Powers Tribunal, later reports indicating the demand was narrowed to United Kingdom users rather than abandoned. The episode rested on a point of mathematics that no statute can repeal: there is no way to give a lawful authority a door into encrypted data without creating a door that an unlawful actor can also use. The Salt Typhoon intrusion of 2024 proved the point in the worst way, when state-sponsored attackers exploited the very lawful-intercept access that telecommunications providers had been compelled to build, reaching the communications of senior officials and even the systems used for court-authorised surveillance. A backdoor for the good guys is, structurally, a backdoor for the bad ones.

There is no mathematics that distinguishes a lawful door into encrypted data from an unlawful one. There is only one door.

India is no exception to the State's reservation of access. Section 69 of the Information Technology Act, 2000, read with the interception, monitoring and decryption rules of 2009, empowers the issuing of decryption directions in defined circumstances, and the DPDP Act expressly exempts instrumentalities of the State, under Section 17(2)(a), from most of its obligations where they process data in the interests of sovereignty, security or friendly relations with foreign States. The DPDP Act, for all that it created genuine rights against private fiduciaries, did not reform the surveillance architecture, and the asymmetry this produces should be stated honestly to any client: in every one of these regimes, data protection is, in the last analysis, protection against everyone except, in defined and contested circumstances, the State itself. The custodian guards the data against the world; the State retains the key to the gate.

X. A working discipline for the large-scale handler

For an enterprise handling personal data at scale, the practical response to all of this is neither paralysis nor box-ticking, but a standing discipline. It begins with knowing what you hold: a genuine data inventory and a record of processing, mapping every category of data, its source, its purpose, its location and its destinations, because you cannot protect, classify or lawfully transfer what you have not mapped. It proceeds to classification, by sensitivity and by destination, so that the most sensitive data and the most exposed corridors receive the most protection. It demands minimisation and time-bound retention, since the data you do not hold cannot be breached, transferred unlawfully, or memorised by a model. It requires a clear lawful basis and a consent architecture that the individual can understand and withdraw.

For cross-border flows it requires a transfer mechanism chosen corridor by corridor: adequacy or Standard Contractual Clauses or Binding Corporate Rules for European outflows, compliance with the negative list and any sectoral localisation for Indian ones, and an eye on the strategic regimes that treat bulk data as a security matter. It requires rigorous contracting with every processor and sub-processor, because liability follows the data into their hands. It requires encryption at rest and in transit, disciplined key management, tokenisation, and, where AI is in use, the privacy-enhancing techniques suited to the risk. It requires a breach-response plan built around a 72-hour clock and a notification path to the regulator and to affected individuals. For those designated as significant fiduciaries it requires periodic impact assessments and independent audits. And it requires that all of this be owned at board level and that the changing law be monitored continuously, the publication of any restricted-country list, the next phase of the DPDP timeline, the AI Act deadline of 02/08/2026, because in this field the law does not stand still and a compliant posture is one that is maintained rather than achieved.

The single most useful operating assumption is the least comfortable one: assume the breach. Design, contract and govern as though the data will one day be exposed and you will one day be asked to account for it, because in a sufficiently long horizon, for a sufficiently large dataset, that day tends to arrive.

A custodian, not an owner

The custodial paradox cannot be resolved, only administered. The data an enterprise holds is valuable precisely because it describes people, and it is for that very reason not freely its own to sell, to expose, or to lose. Sovereignty will go on trying to pin a borderless thing to territory; encryption will go on promising a secrecy it cannot wholly deliver; artificial intelligence will go on enlarging at once the value of data and the difficulty of containing it. The task of the enterprise, and of the advocate who counsels it, is therefore not to escape the paradox but to hold its two halves in honest balance: to treat the data as the asset it commercially is, while guarding it as the trust it legally remains. To move it only where the law permits, to protect it with the best instruments available while admitting their limits, and to remember, in every decision, that one is a custodian and not an owner, accountable in the end for every byte one was permitted to hold.

Watch · Two-minute explainer

India's Data Protection Law

A short video version of this article from the Office's channel, for readers who would rather hear the argument than read it.

Sources and provisions referred to

Statutes and rules

Cases

Official documents

Frequently asked questions

What is the Custodial Paradox in data protection?

It is the contradiction between how business and law view personal data. The enterprise treats data as an asset to own and exploit; the law treats it as a trust held in custody for the person it describes, usable only within the terms of that custody. The two views conflict, and the unmanaged gap between them is where regulatory and civil liability accumulates.

Does India's data protection law apply to companies based abroad?

Yes. The Digital Personal Data Protection Act applies extraterritorially, reaching the processing of the personal data of Indian data subjects wherever in the world that processing takes place. As with the EU's GDPR, the protection attaches to the person, not to the location of the server, so hosting data offshore does not place it beyond the law of the data subject's home.

Is anonymised data safe to use and sell freely?

Far less than is commonly assumed. True anonymisation is fragile, and data that can be re-identified by cross-referencing it against other datasets was never anonymous in law. Misclassifying re-identifiable data as a safe asset converts it into a distributed liability, because every copy across servers and backups carries the obligation the enterprise believed it had removed.

What is the most serious regulatory consequence of a data failure?

Not the headline fine, but the regulatory transfer ban. A fine is paid and absorbed; an order prohibiting the movement of a category of data across a border can halt operations entirely by cutting the enterprise off from its own processing and customers. For a cross-border business, the transfer ban is the existential risk, and the fine is merely the visible one.

Can encryption include a lawful backdoor for the government only?

No. Encryption does not distinguish who holds the key, so a backdoor created for a lawful authority is inherently available to any threat actor who finds it. There is no cryptographic mechanism that admits the State while excluding the attacker, which is why the demand for a backdoor reserved for government rests on a mathematical impossibility.

These answers are educational summaries of the article above; they are not legal advice on any individual matter.

Corrections: if you find an error of law or fact in this article, write to contact@lokeshbhardwajlawoffices.com with the subject 'Correction'. Corrections are made in the text and noted here with the date.

Disclaimer

This article is offered as general legal commentary for informational and educational purposes only and does not constitute legal advice. References to the Digital Personal Data Protection Act, 2023, the European Union's General Data Protection Regulation and Artificial Intelligence Act, sectoral localisation rules including those of the Reserve Bank of India, and connected enactments are to those instruments as in force on the date of writing. The application of data-protection law is fact-specific and jurisdiction-specific. Readers should consult qualified counsel before acting. Published in compliance with the Bar Council of India Rules, Chapter II, Part VI, Rule 36. No part of this article is intended to solicit professional engagement.

WhatsApp