Technology Confronts the Future of Freedom: D.Y. Chandrachud's Dissent in K.S. Puttaswamy v. Union of India on Aadhaar

On 26 September 2018 a five-judge Constitution Bench upheld the Aadhaar Act by 4:1. Justice D.Y. Chandrachud, writing alone, dissented in 339 paragraphs across twelve sections. A reading of his opinion as the first major application of the Puttaswamy proportionality test, by its principal author, to a contested national legislation.

On 26 September 2018 a five-judge Constitution Bench of the Supreme Court upheld, with reading-downs, the Aadhaar (Targeted Delivery of Financial and Other Subsidies, Benefits and Services) Act, 2016 and the broader Aadhaar project. The vote was 4:1. Justice A.K. Sikri spoke for himself, Chief Justice Dipak Misra, and Justice A.M. Khanwilkar; Justice Ashok Bhushan wrote a separate concurring opinion. Justice D.Y. Chandrachud, writing alone, dissented. He would have struck down the entire Aadhaar Act as unconstitutional, directed the destruction of the data already collected under Article 142 in the event Parliament failed to enact a conforming statute within a year, and held that the architecture of mass biometric identification — absent informed consent, absent a robust data-protection regime, absent independent regulatory oversight — was incompatible with the right to privacy the Court had recognised the year before in K.S. Puttaswamy (Retd.) v. Union of India (2017). The dissent runs to 339 paragraphs across twelve sections. It is one of the longest single-judge opinions in the Court’s history.

The author of the dissent was the same judge who, in 2017, had written the plurality opinion in Puttaswamy (I) for himself and three colleagues, establishing privacy as a fundamental right and laying down a four-prong proportionality test as the standard against which intrusions into privacy must be measured. The Aadhaar dissent of 2018 is, in substantial part, the first major application of that test to a contested national legislation by the test’s principal author. The doctrinal stakes of the dissent are therefore unusual: it offers not only a critique of Aadhaar but a template for how the proportionality standard is to operate in Indian constitutional law.

The framing principle of the dissent is set out at paragraph 13. Efficiency is a significant facet of institutional governance. But efficiencies can compromise dignity.¹ When efficiency becomes a universal mantra by which fundamental freedoms are steam-rolled, the dissent observes, there is a danger of a society crossing the line which divides democracy from authoritarian cultures. The proposition is not directed at the Aadhaar project alone. It articulates a constitutional principle for an age in which welfare administration, surveillance architecture, and identification technology operate through a single integrated system. The dissent’s task is to bring that principle into operation through the doctrinal machinery the Court has at its disposal: proportionality, informational privacy, dignity, the protection of constitutional procedure, and the rule of law.

The case the five-judge bench was asked to decide

The petition that gave the case its name was filed by Justice K.S. Puttaswamy, a retired judge of the Karnataka High Court, in 2012. By the time the matter reached the five-judge bench in 2018, more than thirty connected petitions had been consolidated. The petitioners included activists, academics, journalists, retired civil servants, and former judges. The respondents were the Union of India and the Unique Identification Authority of India, the statutory body that administers the Aadhaar project.

The factual scale of the project was undisputed. By the time of the judgment, the Aadhaar database — the Central Identities Data Repository (CIDR) — held the biometric records (fingerprints, iris scans, photographs) and demographic details of approximately 1.2 billion individuals. The dissent describes this at paragraph 10 as creating the largest biometric identity project in the world. The project had begun in 2009 as an executive scheme without statutory backing. The Aadhaar Act of 2016, passed as a Money Bill in the Lok Sabha, gave the project legislative form. Between 2016 and 2018 the use of Aadhaar had expanded from a voluntary identification scheme into a near-mandatory requirement for opening bank accounts, obtaining mobile SIM cards, filing income-tax returns, receiving subsidised foodgrains, accessing pension and welfare benefits, school admissions, and the operation of nearly every interface between the citizen and the state.

The challenges were on multiple grounds. Procedurally, the Act could not have been certified as a Money Bill under Article 110 because its provisions extended well beyond matters relating to expenditure from the Consolidated Fund. Substantively, the architecture of biometric collection, centralised storage, and authentication-based access to benefits violated the right to privacy recognised in Puttaswamy (I). Section 7 (linking welfare benefits to Aadhaar authentication), Section 33 (disclosure on the order of a district judge or in the interest of national security), Section 47 (cognisance only on UIDAI complaint), Section 57 (use of Aadhaar by private body corporates and individuals), and Section 59 (retrospective validation of pre-Act enrolments) were said to be independently unconstitutional. The project as a whole was said to create a surveillance infrastructure incompatible with constitutional democracy.

Section A — the constitutional framing for a technological age

The opening section of the dissent is unusual in its register. Sixteen paragraphs of constitutional theorising precede the doctrinal analysis. The argument is that the constitutional framework was developed in an age that did not contemplate the technologies the case requires the Court to assess. The law evolved instruments of governance in incremental stages, the dissent observes, suited to the social, political and economic context of paper-based administration. The limits of paper allowed for a certain freedom: the freedom of individuality and the liberty of being obscure. Governance with paper could lapse into governance on paper. Technology, the dissent argues, has become a universal language that straddles cultures and tests the limits of the doctrines democracies have evolved to preserve the sanctity of the individual.

The framing draws on Granville Austin’s reading of the Indian Constitution as a seamless web spun from three strands: protecting and enhancing national unity and integrity; establishing the institutions and spirit of democracy; and fostering a social revolution to better the mass of Indians. Each strand depends on the others. A pull on one shakes the balance between them. The challenge the Aadhaar project poses, the dissent suggests, is one of harmonisation: the welfare commitment that motivates the project must be balanced against the protection of liberty and dignity that the Constitution treats as foundational.

The most-quoted line from this section is at paragraph 3. Technology confronts the future of freedom itself. The sentence captures the dissent’s claim that the question before the Court is not a narrow administrative-law question about subsidy delivery but a broader constitutional one about the relationship between the State and the citizen in an information age. The closing of Section A makes the point explicit: the case will determine the future of freedom.

Proportionality and the framework of Puttaswamy (I)

The doctrinal centre of the dissent is its application of the proportionality test laid down in Puttaswamy (I) (2017). The test has four prongs. First, the measure must pursue a legitimate state aim. Second, there must be a rational nexus between the means adopted and the aim. Third, the measure must be the least restrictive means available to achieve the aim — the necessity requirement. Fourth, the means must be proportional in the strict sense: the benefit must outweigh the burden imposed on the protected right.

On the first prong, the dissent accepts that the State has a legitimate aim. Identification of beneficiaries is integral to the fulfilment of social welfare schemes. The targeting of subsidies, the prevention of leakages, and the elimination of duplicate beneficiaries are objects that fall within the constitutional mandate of Articles 38 and 39. Sub-conclusion 6 records this acceptance: there is a legitimate state aim in maintaining a system of identification to ensure that welfare benefits reach those who are entitled.

On the second prong, the dissent accepts that biometric identification can establish a rational nexus with the aim of de-duplication. The criticism the dissent advances is not that the means cannot achieve the aim but that the architecture chosen does more than the aim requires, and produces side-effects the Constitution treats as harms.

The third prong is where the dissent’s engagement is most extended. The State, the dissent observes, had not demonstrated that less intrusive measures would not have served its purposes. The legitimate aim of the State can be fulfilled by adopting less intrusive measures as opposed to the mandatory enforcement of the Aadhaar scheme as the sole repository of identification. The argument turns on the fact that the Aadhaar architecture is not merely a means of identification — it is a means of identification that creates a permanent, centralised, biometric record that can be linked across databases, made available to private body corporates, and used as a bridge across the silos of personal information the Constitution had previously kept separate. The dissent’s observation at sub-conclusion 14(m) is that when Aadhaar is seeded into every database, it becomes a bridge across discreet data silos, which allows anyone with access to this information to re-construct a profile of an individual’s life.

On the fourth prong — the balance — the dissent draws on Granville Austin’s and Upendra Baxi’s engagement with the bread-vs-freedom debate. The dissent’s position is that the framing of the question as a choice between welfare and liberty is a false antithesis. The State must work toward both. The constitutional answer is not to sacrifice one for the other but to design the welfare apparatus so that it does not foreclose the protections of dignity and informational self-determination on which liberty depends.²

Biometric data and what it does to the proportionality calculation

Section F of the dissent develops the proposition that biometric data is constitutionally distinct from other categories of personal information. The reason is technical and the dissent treats it as decisive. Once a biometric system is compromised, it is compromised forever. A password can be changed; a credit card can be reissued; a biometric record cannot. The fingerprint and iris scan that identify an individual at age forty will identify the same individual at age seventy. A breach of biometric data is therefore a breach without a remedy of replacement.

The constitutional implication, the dissent argues, is that the standards for collection, retention, and use of biometric data must be substantially more demanding than for ordinary personal data. At the time of collection, individuals must be informed about the purpose and nature of the collection, who will have access, the retention period, and the rights to access, correct, and delete. The retention period must be justified and bounded. The architecture must include independent regulatory oversight. The Aadhaar Act, the dissent observes, contains none of these protections in adequate form.

The dissent’s critique of specific provisions follows from this premise. Section 28(5) of the Act denies the individual access to her own biometric information — a denial the dissent treats as violative of the fundamental principle that ownership of an individual’s data must at all times vest with the individual. Section 29(4)’s permission for UIDAI to publish or display core biometric information for purposes specified by regulations suffers from overbreadth. The definitions in Section 2 give UIDAI discretionary power to expand the categories of biological information it collects. None of these provisions, the dissent argues, satisfies the heightened standard that biometric data requires.

The architecture of authentication, exclusion, and surveillance

The proportionality analysis would not be complete without considering what the Aadhaar architecture produces in operation. The dissent’s engagement here is grounded in the documentary record. Authentication failures — fingerprint mismatches, iris-scan errors, technical glitches at the point of authentication — had been documented by the Economic Survey of India 2016–17 and by independent research. The consequences fell disproportionately on those whose biometrics had been altered by age, manual labour, or disability. Pensioners with worn fingerprints, manual labourers, the elderly, and the disabled were the groups most likely to be denied benefits at the point of authentication. The dissent records at sub-conclusion 13 that the Aadhaar project has failed to account for and remedy the flaws in its framework and design which has led to serious instances of exclusion of eligible beneficiaries. Dignity and the rights of individuals, the dissent observes, cannot be made to depend on algorithms or probabilities.

Section 7 — which permits the linking of Aadhaar to subsidies, benefits, and services from the Consolidated Fund — is treated by the dissent as the engine of this exclusion. The broad definitions of “services” and “benefits” enable the government to regulate almost every facet of its engagement with citizens. If the requirement of Aadhaar is made mandatory for every benefit or service which the government provides, it is impossible to live in contemporary India without Aadhaar. Section 7 is therefore arbitrary and violative of Article 14.³

Section 57 — the provision permitting body corporates and individuals to use Aadhaar authentication — is treated as constitutionally fatal. The provision creates a pathway by which the biometric architecture of the State is made available to private commercial actors without consent, accountability, or limit. The dissent reads Section 57 as enabling commercial exploitation of personal data, behavioural profiling, and the prediction of preferences. The provision, on the dissent’s analysis, cannot pass constitutional muster.

The Money Bill question and the institutional dimension

Section E of the dissent addresses the procedural challenge. The Aadhaar Act had been certified as a Money Bill by the Speaker of the Lok Sabha and passed by the lower house alone, the Rajya Sabha’s role being limited under Article 109 to non-binding recommendations within fourteen days. The petitioners argued that this had been improper: the Act’s provisions extended well beyond the sub-clauses (a) to (g) of Article 110(1), and the Speaker’s certification had bypassed the bicameral process the Constitution required.

The dissent’s holding is in three parts. First, the Speaker’s certification under Article 110(3) is judicially reviewable. The phrase “shall be final”, the dissent observes, was intended to settle the question between the two Houses of Parliament and as between Parliament and the President; it does not exclude judicial review where substantive illegality or constitutional violation is alleged. The earlier three-judge bench decision in Mohd Saeed Siddiqui and the two-judge bench decision in Yogendra Kumar Jaiswal are overruled. Second, on the merits, the Aadhaar Act fails the Article 110(1) test: provisions other than Section 7 do not fall within sub-clauses (a) to (g), and a Money Bill must contain “only provisions” within those sub-clauses. Third, the proper certification of a Bill as a Money Bill, or otherwise, has implications for the basic structure of the Constitution. The Rajya Sabha’s role in bicameral federalism is part of that basic structure; the Speaker’s improper certification denudes the upper house of its constitutional function.

The Money Bill holding stands independently of the substantive privacy analysis. Even if every provision of the Aadhaar Act had satisfied the proportionality test, the dissent observes, the Act would still be unconstitutional because it had not been validly enacted.⁵

Defiance of interim orders and constitutional morality

Section K is the dissent’s engagement with what it treats as a separate institutional problem. Between 2013 and 2017 the Court had repeatedly ordered, in interim proceedings, that Aadhaar enrolment would remain voluntary and that no person could be denied any service or benefit for want of an Aadhaar number. The Government’s response had been to issue successive notifications and to enact provisions making Aadhaar mandatory for bank accounts, mobile SIM cards, school admissions, pension disbursements, foodgrain rations, and a host of other engagements with the State. The dissent treats this as a defiance of the Court’s interim orders — an institutional posture that, in the dissent’s reading, raises a question of constitutional morality.

The dissent’s observation at sub-conclusion 20 is direct. Defiance of judicial orders (both interim and final) be it by the government or by citizens negates the basis of the rule of law. Propriety and constitutional duty, the dissent observes, required the Union government to move the Court for variation of the interim orders before legislating around them. Institutions of governance are bound by a sense of constitutional morality which requires them to abide by judicial orders.⁶

Identity and what the Constitution protects

The dissent’s closing philosophical claim, at sub-conclusion 21 of paragraph 339, gives the analysis its broadest frame:

Identity is necessarily a plural concept. The Constitution also recognizes a multitude of identities through the plethora of rights that it safeguards. The technology deployed in the Aadhaar scheme reduces different constitutional identities into a single identity of a 12-digit number and infringes the right of an individual to identify herself/himself through a chosen means.D.Y. Chandrachud J., sub-conclusion 21 of para 339

The distinction the dissent draws between identification and identity is the philosophical pivot of the opinion. Aadhaar, in the dissent’s reading, is an instrument of identification — a tool that facilitates proof that the person presenting themselves at a service counter is the person to whom the State’s database had previously assigned a number. Identity, by contrast, is a constitutional concept. The Constitution protects a plurality of identities: identities formed through religion (Article 25), language and culture (Articles 29 and 30), occupation (Article 19(1)(g)), association (Article 19(1)(c)), expression (Article 19(1)(a)), and the autonomy of personal choice that runs through Articles 14 and 21. The dissent’s objection is that an architecture of identification that collapses this plurality into a single twelve-digit number, and that conditions access to citizenship’s practical entitlements on the production of that number, does constitutional damage at the ontological level.⁴

What the dissent makes available

The Aadhaar Act was not struck down. The majority’s reading-downs of Sections 33 and 47, and the striking down of Section 57 to the extent it permitted private use, were the operative outcomes. The Aadhaar project continues. The biometric database remains. The architecture of authentication, with the modifications the majority introduced, remains the legal regime within which Indians interact with the State.

What the dissent makes available, against this backdrop, is a worked constitutional framework that future privacy jurisprudence has to engage. The framework has three components. The first is the application of the proportionality test from Puttaswamy (I) to a contested national legislation by the test’s principal author — a worked example of how the four prongs operate against a real architecture. The second is the doctrinal anchor on biometric data as a constitutionally distinct category requiring heightened protection at every stage from collection through retention to use. The third is the institutional analysis: the Money Bill holding, the basic-structure implication of the Speaker’s certification, and the rule-of-law section on the defiance of interim orders.

The constitutional framework the dissent develops is now part of the body of available reasoning that future benches, in privacy litigation, data-protection litigation, and Money Bill challenges, can draw upon. The closing line of Section A — that the case will determine the future of freedom — placed the stakes at their highest. The dissent’s answer to its own question is on the record, in 339 paragraphs, available to whichever bench in whichever decade next has occasion to consider the constitutional architecture of mass biometric identification.