Public argument about the Aadhaar Act 2016 usually begins from a question the statute does not ask. The question in circulation is whether Aadhaar is mandatory, as if the Act contained a single switch that made the number compulsory for life in general. It does not. The Aadhaar (Targeted Delivery of Financial and Other Subsidies, Benefits and Services) Act, 2016, Act 18 of 2016, is a delivery statute for public money. It creates an identification system, staffs an authority to run it, protects the information the system holds in unusually absolute language, punishes misuse, and then, in one provision, permits the government to demand the number as a condition of receiving certain payments. Everything about the statute’s reach, and nearly everything about the litigation it attracted, follows from the limits built into that one permission.

Stated in terms both camps accept, the design is this. Every resident is entitled to obtain an Aadhaar number by submitting demographic and biometric information. The number is unique to the person, is generated at random, and is not, by itself, proof of citizenship or domicile. The Unique Identification Authority of India enrols residents, maintains the Central Identities Data Repository in which the numbers and the information behind them sit, and performs authentication when an authorised entity asks whether a number and the fingerprint or iris scan presented with it match the record. Core biometric information may not be shared with anyone for any reason whatsoever, and may be used for no purpose other than generating numbers and authenticating them. Over that machinery sits the conditionality provision, Section 7, which allows the Centre or a State to require authentication, or proof of possession of a number, as a condition for receiving a subsidy, benefit or service, but only where the expenditure is incurred from, or the receipt forms part of, the Consolidated Fund.

A reader who holds those elements can answer the practical questions the statute generates, because the Act’s logic is unusually visible. The lawfulness of any demand for Aadhaar turns on how the thing demanded is financed, not on how important it is, not on how convenient it would be for the demanding institution, and not on how many other institutions make the same demand. A cooking gas subsidy may be conditioned on Aadhaar because the money leaves the Consolidated Fund of India. A school desk may not, because admission to a school is an entitlement rather than a funded benefit, and because no expenditure on the child flows from the Fund in the manner Section 7 describes. A mobile telephone connection may not, because a private contract for a commercial service is not a subsidy, benefit or service paid for out of public funds. The Supreme Court’s judgment in the Aadhaar case worked through the statute provision by provision and removed several parts of it, and Parliament amended the Act in 2019 to rebuild private use on consent rather than on the open licence the original text had granted. But the centre of gravity never moved. It sits in Section 7, in a public finance concept that most readers of the Act have never had explained to them.

This article therefore proceeds from the text outward. It sets out what the statute builds, quotes the conditionality provision and the biometric restriction in full, explains the passage of the Act as a Money Bill and the unresolved constitutional question that route left behind, records at headline level what the courts struck down and what they left standing, works through the 2019 amendment change by change, and stages the argument about exclusion and savings at full strength on both sides. The two judgments that reshaped the Act, the nine judge privacy decision of 2017 and the five judge Aadhaar decision of 2018, are treated in depth in the article on the Puttaswamy judgments, which owns that ground in this series. This pillar states what they did to the statute and links to that treatment rather than repeating the reasoning.

The condition is tied to the Fund, not to the service

Section 7 is the operative core of the Act, and the Supreme Court described it in its judgment as the most important provision of the statute, the one around which the entire architecture is built. Its full text, as enacted, is short enough to be read whole, and every word of it carries weight.

The provision reads: “The Central Government or, as the case may be, the State Government may, for the purpose of establishing identity of an individual as a condition for receipt of a subsidy, benefit or service for which the expenditure is incurred from, or the receipt therefrom forms part of, the Consolidated Fund of India, require that such individual undergo authentication, or furnish proof of possession of Aadhaar number or in the case of an individual to whom no Aadhaar number has been assigned, such individual makes an application for enrolment: Provided that if an Aadhaar number is not assigned to an individual, the individual shall be offered alternate and viable means of identification for delivery of the subsidy, benefit or service.”

Five features of that text decide cases, and each is routinely dropped in summaries of the Act.

The power is permissive. The section says a government “may” require authentication or proof of possession. It does not itself impose the requirement on anyone. The obligation on an individual arises only when a government issues a notification under the section for a particular scheme, and the validity of any given demand therefore depends on the notification and on the scheme it covers. A demand for Aadhaar that cannot be traced to a Section 7 notification, or to some other statute that independently authorises it, has no home in this Act.

The trigger is a funding test. The subsidy, benefit or service must be one “for which the expenditure is incurred from, or the receipt therefrom forms part of, the Consolidated Fund of India”. The Consolidated Fund is the Constitution’s central account for each government, the fund into which revenues and loan receipts flow and out of which expenditure is made under appropriations voted by the legislature. Programmes paid from it, which include the public distribution system, the rural employment guarantee, cooking gas subsidies, scholarships and social pensions, fall inside the test. Activities that a government performs without a Fund financed payment to the individual do not, however welfare flavoured they may be. That is why the location of the financing, a bookkeeping fact, became the boundary of a fundamental rights dispute.

The section offers the government three escalating options, not one. It may require the individual to undergo authentication, which is the live matching of the number against biometric or demographic information held in the repository. It may instead require the individual merely to furnish proof of possession of an Aadhaar number, which is a documentary demand satisfiable by producing the number or its letter without any live matching. Or, where no number has been assigned, it may require the individual to apply for enrolment. The third limb matters because it shows the section contemplating a person who has no number at all and treating the answer as enrolment, not exclusion.

The proviso is a right, not a courtesy. Where an Aadhaar number has not been assigned to the individual, the individual “shall be offered alternate and viable means of identification for delivery of the subsidy, benefit or service”. The word “shall” makes the offer obligatory, and the words “alternate and viable” set a quality standard for what must be offered: a means of identification that exists in fact and that the person can actually use. A government that has notified Aadhaar for a scheme and then turns away an unnumbered applicant without offering an alternative route to identification acts against the text of its own enabling provision. The proviso is the statute’s answer, written into the section itself, to the objection that conditionality must exclude those the system has not reached.

The reach is confined to subsidies, benefits and services. The three words are defined broadly in Section 2, a “benefit” being any advantage, gift, reward, relief or payment in cash or kind, and a “service” any provision, facility, utility or other assistance in any form. But breadth of definition does not escape the funding condition, which qualifies all three nouns alike. It is the conjunction of a broad subject list with a narrow funding test that produces the statute’s real shape, and it is the funding test that does the limiting work in every contested application.

The 2019 amendment altered the funding test in one respect, inserting a reference to the Consolidated Fund of the State alongside the Fund of India, so that schemes financed from a State’s own consolidated fund stand on the same footing within the section. It also added a further proviso protecting children, stating that a child shall not be denied any subsidy, benefit or service under the section for failure to establish identity by authentication, by furnishing proof of possession, or, where no number has been assigned, by producing an application for enrolment. Both changes are worked through in the amendment section of this article. The architecture they modify is the architecture described here.

What the Act builds, chapter by chapter

The statute has fifty nine sections arranged in eight chapters, and its structure repays a walk through because the later chapters are where the safeguards and the offences sit, and they are the parts most commentary skips.

Chapter I contains the short title, the extent and the definitions in Section 2. Three definitions do structural work. A “resident” is an individual who has resided in India for a period or periods amounting in all to one hundred and eighty two days or more in the twelve months immediately preceding the date of application for enrolment. Entitlement to a number therefore turns on residence, not citizenship, which is why the Act can serve a population that includes non citizens lawfully present in India while Section 9 simultaneously provides that the number confers no right of citizenship or domicile. “Demographic information” includes name, date of birth, address and other relevant information specified by regulations, and expressly excludes race, religion, caste, tribe, ethnicity, language, records of entitlement, income and medical history. The exclusions are legislative choices about what the identity system will not know about a person, and they are written into the definition itself rather than left to administrative practice. “Core biometric information” means fingerprint, iris scan or such other biological attribute as may be specified by regulations, and it is that subset of the record which the Act wraps in its strictest protection.

Chapter II governs enrolment. Section 3 entitles every resident to obtain an Aadhaar number by submitting demographic and biometric information through the enrolment process, and obliges the enrolling agency to inform the individual at enrolment of the manner in which the information will be used, the nature of the recipients with whom it is intended to be shared during authentication, and the existence of a right to access the information with the procedure for making such a request. Section 4 states the properties of the number: it shall not be re assigned to any other individual, it is a random number bearing no relation to the attributes or identity of its holder, and in physical or electronic form, subject to authentication and other conditions specified by regulations, it may be accepted as proof of identity of the holder for any purpose. Section 5 requires the Authority to take special measures to issue numbers to women, children, senior citizens, persons with disability, unskilled and unorganised workers, nomadic tribes and persons without a permanent dwelling house. Section 6 empowers the Authority to require holders to update demographic and biometric information so that the repository remains accurate. A random number that bears no relation to the holder’s attributes is a deliberate anti profiling device: the digits encode nothing about region, community or date, and cannot be reverse read by anyone who obtains them.

Chapter III governs authentication, and contains Section 7, examined in the previous section, flanked by Section 8 on the mechanics of authentication and Section 9 on the citizenship disclaimer. Section 10 provides for the Central Identities Data Repository, the centralised database in one or more locations containing all Aadhaar numbers with the corresponding demographic and biometric information, which the Authority may engage one or more entities to establish and maintain.

Chapter IV establishes the Authority and is treated in the next section of this article. Chapter V deals with grants, accounts and audit, including the provision that the accounts of the Authority are audited by the Comptroller and Auditor General of India and that its annual report is laid before each House of Parliament. Chapter VI, Sections 28 to 33, is the protection code: security and confidentiality of information, the restriction on sharing, the treatment of biometric information as sensitive personal data, alteration of records, access by the holder to his or her own authentication record, and the two narrow disclosure gateways. Chapter VII, Sections 34 to 47, creates the offences and penalties and restricts cognizance of them. Chapter VIII contains the miscellaneous provisions, among them the power of the Central Government to supersede the Authority, the rule making and regulation making powers, and the savings provision in Section 59 which validates things done under the executive notifications that preceded the Act.

Two general provisions deserve notice because they answer questions readers bring to the statute from outside it. Section 56 provides that the Act is in addition to, and not in derogation of, any other law in force, which is why the Aadhaar Act does not by itself settle what the tax law or the money laundering law may separately demand. Section 44 gives the Act extraterritorial reach over offences involving data in the repository, and Section 45 provides that offences under the Act are to be investigated by a police officer not below the rank of Inspector.

The Authority, its money, and who can remove it

The Unique Identification Authority of India is the institutional centre of the Act. Section 11 requires the Central Government, by notification, to establish the Authority, responsible for the processes of enrolment and authentication and for the other functions the Act assigns. It is a body corporate with perpetual succession and a common seal, headquartered in New Delhi, able to establish offices elsewhere in India with the prior approval of the Central Government. Section 22 transferred to it, on establishment, all the assets and liabilities of the executive era Authority created by the Planning Commission notification of 28 January 2009, together with all data and information collected during enrolment and all details of authentication performed under the old arrangement, so that the statutory body inherited the programme whole rather than building it afresh.

How is the Unique Identification Authority of India constituted?

It consists of a Chairperson, appointed part time or full time, two part time Members, and a chief executive officer who serves as Member Secretary, all appointed by the Central Government. The Chairperson and Members must be persons of ability and integrity with at least ten years of experience in fields such as technology, governance, law, development, economics or administration.

The chief executive officer, who must not be below the rank of an Additional Secretary to the Government of India, is the legal representative of the Authority and is responsible for its day to day administration, for implementing its work programmes and decisions, for preparing its statement of revenue and expenditure and executing its budget, and for submitting an annual general report, programmes of work, annual accounts and the coming year’s budget to the Authority for approval. The Chairperson and Members hold office for a term of three years, are eligible for reappointment, and may not hold office after attaining the age of sixty five. Removal before the term ends is confined by Section 15 to specified grounds, including insolvency, physical or mental incapacity, conviction of an offence involving moral turpitude, acquisition of a prejudicial financial or other interest, and abuse of position rendering continuance detrimental to the public interest, with a reasonable opportunity of being heard required for several of those grounds. Section 16 then imposes a cooling off regime: a former Chairperson or Member may not, without the previous approval of the Central Government, accept employment with an entity associated with the Authority’s work for three years, act for any person in a proceeding to which the Authority is a party and in which the former office holder acted for the Authority, advise any person using information obtained in office, or take up service with an entity with which the office holder had direct and significant official dealings during the term.

The functions conferred by Section 23 are extensive and are exercised largely through regulations. The Authority develops the policy, procedure and systems for issuing Aadhaar numbers and performing authentication. It specifies the demographic and biometric information required for enrolment and the processes for collecting and verifying it, appoints the entities that operate the repository, generates and assigns the numbers, performs authentication, maintains and updates the information in the repository, and may omit and deactivate a number and the information relating to it in the manner specified by regulations. It specifies the manner of use of Aadhaar numbers for providing or availing of subsidies, benefits, services and other purposes, sets the terms for appointing Registrars, enrolling agencies and service providers, calls for information and records and conducts inspections, inquiries and audit of the repository and of the agencies it appoints, specifies processes for data management and security protocols, levies and collects fees, sets up facilitation centres and a grievance redressal mechanism for individuals, Registrars, enrolling agencies and service providers, and promotes research in biometrics and related areas. The Authority may enter into memoranda of understanding with the Central Government, State Governments, Union territories and other agencies, and may appoint Registrars and engage agencies to collect, store, secure and process information.

How are the Authority’s accounts audited, and to whom does it report?

Its accounts are audited annually by the Comptroller and Auditor General of India, whose certified accounts and audit report are forwarded to the Central Government and laid before each House of Parliament. The Authority must also prepare an annual report describing its activities and accounts and its programmes of work for the coming year.

The money provisions complete the accountability design. Under Section 24 the Central Government may make grants to the Authority after due appropriation by Parliament. As enacted, Section 25 provided that the fees and revenues collected by the Authority shall be credited to the Consolidated Fund of India, a provision the 2019 amendment replaced with a dedicated fund for the Authority, as the amendment section of this article records. The Central Government holds two further levers. Under Section 50, the Authority is bound by directions on questions of policy given by the Central Government in writing, though it must as far as practicable be given an opportunity to express its views first, and the section does not empower directions on technical or administrative matters. Under Section 48, the Government may supersede the Authority in specified circumstances.

How can the Central Government take over the Authority, and for how long?

It may supersede the Authority by notification for a period not exceeding six months, where the Authority cannot discharge its functions for reasons beyond its control, has persistently defaulted on government directions or its duties, or a public emergency exists, after giving the Authority a reasonable opportunity to make representations against the supersession.

During supersession the Chairperson and Members vacate office, the powers and duties pass to the persons appointed for the purpose, and the Authority’s properties vest in the Central Government, with reconstitution required on or before the expiry of the period. The notification and a full report of the action taken and the circumstances leading to it must be laid before each House of Parliament at the earliest. The design is the standard Indian regulatory compromise: a specialised statutory body with defined membership, security of tenure qualified by removal grounds, public money channelled through parliamentary appropriation and audit, and a government override that is time limited, reasoned and reportable.

The programme before the statute

The Act of 2016 arrived seven years into the life of the thing it regulated. The Unique Identification Authority of India was created by a notification of the Planning Commission, number A-43011/02/2009-Admin. I, dated 28 January 2009, as an attached office charged with issuing a unique identification number to residents of India. Nandan Nilekani, a co founder of Infosys, was appointed its first Chairperson in 2009 with the rank of a Cabinet Minister, an appointment that signalled the programme’s intended scale from the outset. The first Aadhaar number was issued on 29 September 2010 at Tembhli, a village in the Nandurbar district of Maharashtra, and enrolment expanded through Registrars, principally State Governments, public sector banks and other public agencies, whose enrolling agencies captured names, dates of birth, addresses, photographs, ten fingerprints and two iris scans at enrolment camps and centres across the country.

Why did the identification programme run for seven years before Parliament legislated?

Because it began as an executive scheme under a Planning Commission notification rather than a statute, and the first attempt at a legislative basis, the National Identification Authority of India Bill of 2010, was rejected by the Standing Committee on Finance in 2011 and never enacted, leaving the programme on executive authority until 2016.

The 2010 Bill would have created a National Identification Authority with a structure similar to the one the 2016 Act eventually adopted. The Standing Committee on Finance, chaired by Yashwant Sinha, examined it and reported in December 2011 in terms that were unusually blunt for a committee report: it found the Bill unacceptable, questioned the programme’s direction and cost, noted overlaps and the absence of a settled data protection framework, and recommended that the government reconsider. The Bill was not revived. Meanwhile, the enrolment machinery kept running. Government notifications under other statutes began attaching the number to particular schemes, the courts were moved by petitions challenging the programme as a whole, and by April 2016 the Authority announced that enrolment had crossed one hundred crore residents, a scale that made the absence of a statutory basis a problem of a different order than it had been in 2011. A database of the fingerprints and iris scans of a substantial fraction of the population existed, was being used daily to authenticate transactions, and rested, as a matter of law, on a notification issued by an executive body in 2009.

That history explains two features of the 2016 Act that otherwise look odd. The first is Section 22, which transferred the old Authority’s assets, liabilities, data and contracts to the new statutory Authority, so that the enrolments already made and the authentications already performed passed into the statute’s custody without a break. The second is Section 59, the savings provision, which provides that anything done or any action taken by the Central Government under the resolution embodied in the 2009 Planning Commission notification, or by the Department of Electronics and Information Technology under the Cabinet Secretariat notification number S.O. 2492(E) dated 12 September 2015, shall be deemed to have been validly done or taken under the Act. A validating provision of this kind is the standard instrument by which a legislature adopts the acts of a predecessor executive regime, and its validity was among the questions the Supreme Court later answered in the government’s favour at headline level: the Court upheld Section 59 and treated the enrolment drive of the executive years as covered by it.

The route by which the Bill finally passed in 2016 is treated separately in this article, because it generated a constitutional controversy of its own that outlasted the judgment. What matters for the statute’s history is the sequence. The Aadhaar Bill was introduced in the Lok Sabha on 3 March 2016 by the Finance Minister, Arun Jaitley, certified by the Speaker as a Money Bill. The Lok Sabha passed it on 11 March 2016. The Rajya Sabha returned it on 16 March 2016 with recommendations for amendments, which the Lok Sabha considered and did not accept on the same day. The President assented on 25 March 2016, and the Act was published as Act 18 of 2016. Commencement was staggered by notification under Section 1(3): Sections 11 to 20, 22, 23 and 48 to 59, the provisions constituting the Authority and validating the past, came into force on 12 July 2016, and Sections 1 to 10 and 24 to 47, the enrolment, authentication, finance, protection and offences chapters, came into force on 12 September 2016.

How does the enrolment process treat a person whose fingerprints cannot be captured?

The enrolment framework provides for biometric exceptions. Where a person cannot give some or all of the biometric inputs, because of missing fingers or eyes, ridges worn by manual labour, injury or disability, the exception is recorded during enrolment and the number is issued on what can be captured, rather than the applicant being refused.

The exception machinery matters beyond enrolment. A person enrolled with exceptions, or whose fingerprints degrade with age and labour, is precisely the person most likely to meet a failed authentication at a ration shop terminal, and the Act’s answer to that encounter is not the exception process but the proviso to Section 7 and the exception mechanisms the delivery systems are required to operate. The connection between enrolment quality, authentication reliability and the proviso is where the exclusion debate, staged later in this article, gets its force.

Authentication: what happens, and what the holder is owed

Authentication is defined in Section 2 as the process by which the Aadhaar number along with demographic information or biometric information of an individual is submitted to the Central Identities Data Repository for verification, and the repository verifies the correctness, or the lack of it, on the basis of the information available with it. The definition is worth dwelling on, because it states what authentication is not. The repository answers a matching question about a submitted claim. It does not certify that the person is a citizen, that the person is entitled to anything, or that the transaction in aid of which the check is run is proper. Those judgments remain with the requesting entity and the scheme. Section 8 adds that the Authority shall respond to an authentication query with a positive, negative or other appropriate response, sharing such identity information as is permitted and excluding any core biometric information: the fingerprint in the repository never travels back down the wire.

The obligations around the transaction sit on the requesting entity, the agency or person that submits the number and information to the repository. Section 8(2) requires the requesting entity, unless otherwise provided in the Act, to obtain the consent of the individual before collecting his or her identity information for authentication, and to ensure that the identity information is used only for submission to the repository for authentication. Section 8(3) requires it to inform the individual of the nature of information that may be shared upon authentication, of the uses to which the information received during authentication may be put by the requesting entity, and of alternatives to submission of identity information to the requesting entity. Section 29(3) adds that no identity information available with a requesting entity shall be used for any purpose other than that specified to the individual at the time of submission, or disclosed further except with the prior consent of the individual. The transaction is therefore consent based, purpose bound and disclosure restricted by the text of the Act, whatever criticisms may be made of practice under it.

Chapter VI then surrounds the record itself. Section 28 obliges the Authority to ensure the security of identity information and authentication records, to take all necessary measures to protect information in its possession or control against access, use or disclosure not permitted under the Act and against accidental or intentional destruction, loss or damage, and to impose equivalent obligations by contract on the agencies, consultants and advisers it engages. Section 28(5) bars the Authority, its officers and any agency maintaining the repository from revealing information stored in the repository or any authentication record to anyone, during service or thereafter, with one exception: the holder may request access to his or her own identity information, excluding core biometric information, in the manner specified by regulations. Section 31 permits a holder to request alteration of demographic or biometric information that is incorrect or has changed, and forbids alteration of identity information in the repository except in the manner the Act or regulations provide. Section 32 requires the Authority to maintain authentication records in the manner and for the period specified by regulations, entitles every holder to obtain his or her authentication record, and then states a limitation of real architectural significance: the Authority shall not, either by itself or through any entity under its control, collect, keep or maintain any information about the purpose of authentication. The repository records that a match was requested by a named requesting entity at a time and answered in a particular way. It is forbidden from building, out of those records, an account of why the citizen transacted.

The strictest sentence in the Act is in Section 29(1), and it deserves quotation in full: no core biometric information, collected or created under this Act, shall be shared with anyone for any reason whatsoever, or used for any purpose other than generation of Aadhaar numbers and authentication under this Act. The provision is drafted without exception or qualification. It does not say shared without consent, or shared without lawful authority, or shared for a sufficient reason. The sharing ban covers everyone and every reason, and the use ban confines the biometric core of the system to exactly two functions. Section 29(2) deals with the remainder of the identity information, providing that demographic and other non core information may be shared only in accordance with the Act and regulations. Section 29(4) forbids the publication, display or public posting of an Aadhaar number or core biometric information except for purposes specified by regulations. Section 30 deems biometric information collected and stored in electronic form to be an electronic record and sensitive personal data or information, and applies the Information Technology Act, 2000 and its rules to it in addition to the Aadhaar Act’s own provisions.

Against that enclosure, Section 33 opens exactly two gateways, and their original form explains why one of them did not survive judicial scrutiny. The first gateway is a court order: nothing in the confidentiality provisions applies to a disclosure of information, including identity information or authentication records, made pursuant to an order of a court not inferior to that of a District Judge, provided the order is not made without giving the Authority an opportunity of hearing. The second gateway is national security: a disclosure made in the interest of national security in pursuance of a direction of an officer not below the rank of Joint Secretary to the Government of India, specially authorised by an order of the Central Government, is likewise permitted, provided every such direction is reviewed by an Oversight Committee consisting of the Cabinet Secretary and the Secretaries to the Government of India in the Department of Legal Affairs and the Department of Electronics and Information Technology before it takes effect, and provided any direction remains valid for three months and may be extended for a further three months after a further review. The 2019 amendment rebuilt both gateways, raising the level of the court and of the officer and adding the holder’s own right to be heard, as the amendment section of this article sets out.

Offences, penalties and the complaint gate

Chapter VII constructs a criminal code for the identity system, and its penalties are calibrated by proximity to the repository. At the outer ring, Section 34 punishes impersonation at the time of enrolment, the giving of false demographic or biometric information while pretending to be another person, dead or alive, real or imaginary, with imprisonment up to three years or a fine up to ten thousand rupees or both. Section 35 applies the same range to changing, or attempting to change, a holder’s demographic or biometric information by impersonation with the intention of causing harm or mischief or appropriating the holder’s identity. Section 36 punishes anyone who, not being authorised to collect identity information, pretends by words, conduct or demeanour to be so authorised, with imprisonment up to three years or a fine up to ten thousand rupees, or in the case of a company a fine up to one lakh rupees, or both.

Section 37 addresses disclosure: intentionally disclosing, transmitting, copying or otherwise disseminating identity information collected in the course of enrolment or authentication to an unauthorised person, or in contravention of an agreement under the Act, is punishable with imprisonment up to three years or a fine up to ten thousand rupees, or in the case of a company one lakh rupees, or both. Section 40 punishes a requesting entity that uses identity information in contravention of the duties in Section 8(3) with imprisonment up to three years or the same fine pattern, and Section 41 punishes an enrolling agency or requesting entity that fails to comply with the intimation requirements of Section 3(2) or Section 8(3) with imprisonment up to one year or a fine up to ten thousand rupees, or one lakh rupees for a company, or both.

At the centre, Section 38 lists the acts that attack the repository itself: intentionally accessing or securing access to the Central Identities Data Repository without authority, downloading, copying or extracting its data, introducing a virus or other computer contaminant, damaging its data, disrupting access to it, denying access to an authorised person, revealing information in contravention of Section 28(5) or sharing, using or displaying information in contravention of Section 29, destroying, deleting or altering stored information, or stealing, concealing, destroying or altering the computer source code used by the Authority with the intention of causing damage. The punishment is imprisonment up to three years and a fine that shall not be less than ten lakh rupees, the mandatory minimum fine marking the legislature’s assessment of the gravity of striking at the central record. Section 39 punishes tampering with data in the repository or in removable storage media, with the intent of modifying information relating to a holder or discovering any information in it, with imprisonment up to three years and a fine up to ten thousand rupees. Section 42 supplies a general penalty, imprisonment up to one year or a fine up to twenty five thousand rupees, or one lakh rupees for a company, for offences under the Act, rules or regulations for which no specific penalty is provided. Section 43 deems every person in charge of, and responsible for, the conduct of a company’s business guilty of an offence committed by the company unless the person proves it was committed without his or her knowledge or that all due diligence was exercised to prevent it, and extends liability to directors, managers, secretaries and other officers where the offence occurred with their consent or connivance or is attributable to their neglect. Section 46 preserves other punishments under other laws, so an Aadhaar Act penalty does not displace prosecution under, for example, the information technology or penal law for the same conduct.

The gatekeeping provision is Section 47. As enacted, it provided that no court shall take cognizance of any offence punishable under the Act save on a complaint made by the Authority or any officer or person authorised by it, and that no court inferior to a Chief Metropolitan Magistrate or a Chief Judicial Magistrate shall try any offence under the Act. The design concentrated the prosecutorial trigger in the Authority itself. Its weakness is visible on its face: a citizen whose information had been misused by an enrolling agency, or disclosed by an insider, could not set the criminal law in motion at all unless the Authority chose to complain, even though the Authority might itself be the institution whose systems or agents were implicated. That defect is among the matters the Supreme Court addressed, and the correction, first directed by the Court and then enacted by Parliament in 2019, opened the complaint to the individual as well, as the judgment and amendment sections of this article record.

The Money Bill route and the question it left open

The Aadhaar Act was passed as a Money Bill, and the mechanics of that classification explain why the passage took the shape it did. Article 110 of the Constitution defines a Money Bill as a Bill containing only provisions dealing with the matters listed in its clauses, which include the imposition and regulation of taxes, the borrowing of money by the Government of India, the custody of and payments into and withdrawals from the Consolidated Fund and the Contingency Fund, the appropriation of moneys out of the Consolidated Fund, expenditure charged on the Fund, the receipt, custody and audit of public moneys, and any matter incidental to those subjects. Under Article 109, a Money Bill may be introduced only in the Lok Sabha. The Rajya Sabha may return it with recommendations within fourteen days, but the Lok Sabha may accept or reject those recommendations as it pleases, and the Bill is then deemed passed by both Houses in the form in which the Lok Sabha leaves it. The Speaker’s decision on whether a Bill is a Money Bill is recorded in a certificate endorsed on the Bill under Article 110(3) and (4).

The Aadhaar Bill received that certificate. The government’s justification, inside and outside the House, was that the statute’s operative provision dealt with expenditure from the Consolidated Fund of India: subsidies, benefits and services funded from the Fund were its subject matter, moneys were appropriated for them, and the identification machinery was the means of delivering that expenditure to the persons entitled to it, which brought the machinery within the incidental clause of Article 110. The opposition case was that the certificate converted a general identification law, touching enrolment of the entire resident population, biometric databases, private sector use and criminal offences, into a measure the Council of States could only advise upon, and that the Fund connection attached only to a fraction of the Bill’s provisions. The Rajya Sabha in fact returned the Bill with recommendations, including substantial ones, on 16 March 2016, and the Lok Sabha declined to accept them on the same day, which the Money Bill procedure permitted it to do without further reference back. The general account of how the Money Bill route differs from an ordinary or financial Bill, and why the certificate matters to the federal balance, is set out in the comparison of Money Bills and Finance Bills.

The challenge to the certificate was decided in the Aadhaar judgment. A majority of the five judge bench upheld the passage as a Money Bill, reasoning that the provisions of the Act in pith and substance fell within Article 110, with Section 7 at the centre and the remaining provisions incidental to it, and holding that the Speaker’s decision, though open to judicial review, disclosed no ground for interference. Justice Chandrachud, dissenting, held that the Money Bill route had been misused: the Act, in his analysis, could not be characterised as containing only the matters Article 110 lists, and passing it by that route debased a constitutional process and amounted, in his phrase, to a fraud on the Constitution. The Court dismissed review petitions against the judgment by a majority of four to one in January 2021, again with a dissent that would have kept the question open. In the meantime, a coordinate bench of five judges in Rojer Mathew v South Indian Bank Limited, decided on 13 November 2019 in a challenge to provisions of the Finance Act, 2017 concerning tribunals, held that the Aadhaar majority had not fully delineated the scope of Article 110, in particular the effect of the word “only” in Article 110(1), and referred the Money Bill questions to a larger bench of seven judges for an authoritative determination. That reference has not been decided. The consequence is an unusual state of affairs: the Aadhaar Act stands, its passage has been upheld by the bench that heard it, and the reasoning by which it was upheld has been expressly doubted by a bench of equal strength and sent upward for reconsideration, where the matter rests.

What the Supreme Court removed, and what it left standing

The Aadhaar judgment, delivered on 26 September 2018 in Justice K.S. Puttaswamy (Retd.) v Union of India, reported at (2019) 1 SCC 1, upheld the Act in its main structure and cut specific provisions out of it. The majority opinion was delivered by Justice A.K. Sikri for himself, Chief Justice Dipak Misra and Justice A.M. Khanwilkar, with Justice Ashok Bhushan concurring in a separate opinion and Justice D.Y. Chandrachud dissenting. The reasoning on privacy, proportionality and surveillance belongs to the judgments and is owned in this series by the treatment of the Puttaswamy cases linked earlier in this article; the concern here is the resulting state of the statute book.

The Court upheld Section 7. It held that the requirement of authentication or proof of possession for subsidies, benefits and services funded from the Consolidated Fund was constitutionally permissible, and it declined to strike the section down on account of instances in which authentication failure had led to denial of a benefit, while insisting on the availability of the exception mechanisms and the alternative means the framework provides. It upheld Section 29, the sharing restriction, as a protection rather than a vice. It upheld Section 139AA of the Income tax Act, 1961, which requires the quoting of Aadhaar in the application for a permanent account number and in the return of income. It upheld Section 59, the savings provision validating what had been done under the 2009 and 2015 notifications.

What it removed was specific. Section 57, the provision that nothing in the Act prevented the use of the Aadhaar number for establishing identity for any purpose, whether by the State or any body corporate or person, pursuant to any law or any contract to that effect, was struck down in the part that enabled private parties to build identification requirements on it: the majority held unconstitutional the use of Aadhaar by any body corporate or person pursuant to a contract, and the enabling words that supported such use. Section 33(2), the national security disclosure gateway at Joint Secretary level, was struck down. Section 47 was struck down in its restrictive effect and the Court directed that the individual affected be enabled to file a complaint, rather than the Authority alone controlling cognizance of offences under the Act. Section 33(1) was read down: the disclosure order of a court must carry safeguards beyond those the text stated, including a hearing to the individual whose information is sought, and the amending Parliament subsequently raised the level of the court itself. Regulation 27 of the Aadhaar (Authentication) Regulations, 2016, which permitted authentication records to be archived for five years, was struck down, the Court holding that retention of authentication records for a period longer than six months is not permissible.

The Court also fixed the boundaries of compulsion outside Section 7, and these holdings are the ones with daily consequences for readers. Aadhaar could not be made mandatory for admission to schools: admission, the majority reasoned, is neither a service nor a benefit within the meaning that matters, a child is entitled to elementary education by right, and the benefit logic of the section does not reach the school gate. It could not be made mandatory for examinations conducted by bodies such as the Central Board of Secondary Education, the National Eligibility cum Entrance Test or the University Grants Commission, which are likewise not subsidies, benefits or services funded in the Section 7 sense. The requirement of linking Aadhaar with mobile telephone connections, which rested on a circular of the Department of Telecommunications dated 23 March 2017 issued in purported compliance with an earlier order of the Court in a different matter, was set aside as unconstitutional. The requirement of linking Aadhaar with bank accounts, which rested on an amendment to the prevention of money laundering rules made in 2017, did not survive the majority’s scrutiny either: compulsory seeding of bank accounts with Aadhaar lacked a sufficient statutory basis and failed the proportionality enquiry. The permanent account number linkage stood on a different footing because Parliament had enacted it in Section 139AA itself, and the Court upheld that provision. The pattern across these holdings is the pattern of the statute: compulsion survived exactly where a legislature had tied it to Fund financed delivery or enacted it directly, and failed where it rested on administrative circulars, subordinate rules or the open private use licence of Section 57.

The 2019 amendment: private use rebuilt on consent

The Aadhaar and Other Laws (Amendment) Act, 2019, Act 14 of 2019, which came into force on 25 July 2019, was Parliament’s response to the judgment. Its legislative path was itself instructive. An amending Bill carrying substantially the same scheme was introduced in December 2018 and passed by the Lok Sabha on 4 January 2019, but was not taken up by the Rajya Sabha before the House was dissolved, and lapsed. The substance was then carried by an Ordinance promulgated on 2 March 2019, and the 2019 Act replaced the Ordinance with the authority of both Houses. The changes fall into five groups.

The first and largest group rebuilds the use of Aadhaar outside Section 7 on voluntariness. The Act substituted the proof of identity provision so that an Aadhaar number holder may voluntarily use the Aadhaar number, in physical or electronic form, by way of authentication or offline verification, with the informed consent of the holder. It defined the two modes distinctly: authentication remains the live check against the repository, while offline verification, inserted as Section 8A, is the process of verifying the identity of the holder through modes specified by the Authority without an authentication transaction, in practice the digitally signed records and codes the Authority issues for that purpose. Offline verification carries its own discipline. The entity seeking it must obtain the consent of the individual, or of a parent or guardian where the individual is a child, must inform the individual of alternatives to sharing information, must use what is collected only for the verification purpose, and is prohibited from collecting, using or storing the Aadhaar number or biometric information of the individual for any purpose. An amendment also recognised the use of an alternative virtual identity, so that a holder could transact by means of a substitute identity rather than exposing the twelve digit number itself.

The second group controls who may authenticate at all. An entity may be allowed to perform authentication only where the Authority is satisfied that the entity complies with the standards of privacy and security specified by the Authority, and where the authentication is either permitted under a law made by Parliament or is prescribed by the Central Government to be in the interest of the State. The open licence of the old Section 57 was replaced by a gated permission, and the gate has two keys, regulatory compliance and legislative or expressly prescribed authority. Consistently with the judgment, the 2019 Act omitted Section 57 altogether. The conforming amendments to the Indian Telegraph Act, 1885 and the Prevention of Money Laundering Act, 2002 then reopened the voluntary route for the two sectors that had lost their compulsory one: a person seeking a mobile connection or a bank account may choose to establish identity by Aadhaar authentication, by offline verification, by passport or by another officially valid document, the choice among the modes is the person’s own, the other modes must be made available, and no person is to be denied a service for not having an Aadhaar number.

The third group strengthened the individual’s position inside the Act. Section 47 was substituted so that a court may take cognizance of an offence on a complaint made by an individual as well as by the Authority, ending the Authority’s monopoly over prosecution. A child who had been enrolled could, on attaining the age of eighteen years, apply for cancellation of the Aadhaar number within six months of attaining majority, converting the consent deficit of childhood enrolment into an adult decision. Section 7 was amended to extend the funding test to the Consolidated Fund of the State and to add the proviso that no child shall be denied any subsidy, benefit or service under the section for failure to establish identity by authentication or proof of possession or by producing an application for enrolment.

The fourth group rebuilt the disclosure gateways of Section 33 along the lines the Court had indicated. In Section 33(1), the court whose order can authorise disclosure was raised from a court not inferior to that of a District Judge to a Judge of a High Court, the opportunity of hearing was extended to the concerned Aadhaar number holder in addition to the Authority, and a new proviso barred the disclosure of core biometric information under the sub section altogether. In Section 33(2), the officer whose direction can authorise disclosure in the interest of national security was raised from Joint Secretary to Secretary to the Government of India, with the Oversight Committee review and the three month validity, extendable by three months, retained.

The fifth group gave the Authority regulatory teeth and its own money. A new Chapter VIA, Sections 33A to 33E, created a regime of civil penalties: where an entity in the Aadhaar ecosystem fails to comply with the Act, the rules or regulations, or directions issued by the Authority under the new Section 23A, or fails to furnish information, documents or returns required by the Authority, it is liable to a civil penalty which may extend to one crore rupees for each contravention, with an additional penalty of up to ten lakh rupees for every day a continuing failure persists after the first contravention, and unpaid penalties are recoverable as arrears of land revenue. Penalties are adjudicated by an Adjudicating Officer, an officer of the Authority not below the rank of Joint Secretary to the Government of India appointed for the purpose, acting only on a complaint made by the Authority and after giving the entity an opportunity of being heard, with an appeal lying to the Telecom Disputes Settlement and Appellate Tribunal within forty five days of the order. The expression “Aadhaar ecosystem” was defined to include enrolling agencies, Registrars, requesting entities, offline verification seeking entities and other entities specified by regulations, and Section 23A armed the Authority to issue directions to them. The money provisions were recast so that the fees, grants and charges received by the Authority are credited to a Unique Identification Authority of India Fund and applied to its expenses, replacing the original direction that fees and revenues be credited to the Consolidated Fund of India.

Where Aadhaar can and cannot be required

The table below is the practical yield of everything the statute, the judgment and the amendment together establish. It states, for each situation a reader is likely to meet, whether authentication or proof of possession of an Aadhaar number may lawfully be required, the source of that rule, and the alternative the person must be offered where one exists. It states the law as enacted, amended and read by the Court; it is legal information, not advice on any person’s case.

Situation	May Aadhaar be required?	Source of the rule	Alternative that must be offered
A subsidy, benefit or service funded from the Consolidated Fund	Yes, authentication or proof of possession may be made a condition by notification	Section 7 of the Aadhaar Act, upheld by the Supreme Court	Alternate and viable means of identification must be offered where no Aadhaar number has been assigned, under the proviso to Section 7
Income tax filing and permanent account number	Yes, the Aadhaar number or enrolment application must be quoted	Section 139AA of the Income tax Act, 1961, upheld by the Supreme Court	The enrolment identification number may be quoted where a number has not been assigned; there is no general opt out
A bank account	No, compulsory linking was struck down; voluntary use is permitted for customer verification	The Aadhaar judgment and the 2019 amendment of the Prevention of Money Laundering Act, 2002	Any other officially valid document, such as a passport; no service may be denied for not having an Aadhaar number
A mobile telephone connection	No, compulsory linking was struck down; voluntary use is permitted for subscriber verification	The Aadhaar judgment setting aside the Department of Telecommunications circular of 23 March 2017, and the 2019 amendment of the Indian Telegraph Act, 1885	Other modes of identification offered by the licensee, which must be made available alongside the Aadhaar modes
Admission to a school	No	The Aadhaar judgment: admission is not a subsidy, benefit or service within Section 7, and elementary education is an entitlement	None is needed, because the demand itself is not authorised
An examination conducted by a board or public body	No	The Aadhaar judgment on examinations conducted by bodies such as the CBSE, NEET and UGC	The candidate identification documents the examining body otherwise prescribes
A private service offered by a company or person	No general requirement is permitted; authentication is possible only through the gated route	Section 57 struck down in its private use part and then omitted; the 2019 consent based framework governs	Any other identity document the provider accepts; Aadhaar may be offered voluntarily by the holder
Employment	No for private hiring as a statutory condition; yes inside the rural employment guarantee, where wages are Fund financed	The 2019 framework for private entities, and Section 7 for Fund financed employment programmes	For private hiring, other identity and address documents; for the employment guarantee, the proviso to Section 7 protection applies

The table’s logic is the article’s argument in miniature. Read down the source column and a pattern appears that no summary of the Act conveys: every “yes” rests on an Act of Parliament and on the funding test, and every “no” rests on the absence of one or the other. Administrative enthusiasm, however widespread, is not in the source column anywhere.

The two cases, at full strength

The case for the statute begins with the condition of welfare delivery before it. India’s subsidy and benefit systems ran, and in large part still run, through paper records maintained at the point of delivery: ration cards, muster rolls, beneficiary lists compiled by local officials. The documented pathology of that system was the duplicate and the ghost: names that appeared twice, names of persons who had died or never existed, entitlements drawn against them, and the value diverted before it reached any real claimant. The Union Government’s case, in Parliament and before the Supreme Court, was that a single, biometrically anchored identity, verified at the point of delivery, attacks exactly that pathology, because a fingerprint cannot be duplicated in the way a name on a list can. The government’s estimate, stated in Parliament in 2018, placed the cumulative savings from Direct Benefit Transfer, the payment architecture that rides on identified bank accounts and seeded beneficiary lists, at more than ninety thousand crore rupees by March 2018, from the deletion of duplicate, fake and ineligible beneficiaries across schemes including the public distribution system, the rural employment guarantee and cooking gas subsidy. The Economic Survey for 2016 to 2017 had earlier credited the reformed delivery of the cooking gas subsidy with a significant reduction in leakage, as subsidised cylinders followed identified consumers rather than anonymous demand. Supporters add an inclusion argument that is often missed by the statute’s critics. The person worst served by the paper system is the person with no documents at all: the migrant worker whose ration card belongs to another State, the widow whose name was never entered, the labourer whose employer holds his papers. A portable identity that a person carries in his or her own body, and that a delivery point anywhere in the country can verify, is on this account a transfer of power from the record keeper to the claimant, and Section 5’s list of the persons for whom the Authority must take special measures, women, children, senior citizens, persons with disability, unskilled and unorganised workers, nomadic tribes and the homeless, is the statute’s recognition of exactly that population.

The case against begins at the same delivery point and reads the same transaction from the other side. Conditioning a subsistence benefit on a technology means that when the technology fails, the benefit fails, and the failure falls on the person least able to absorb it. Fingerprints wear down with manual labour and age. Connectivity at remote delivery points is unreliable. Seeding errors, a wrongly linked number, a name spelled differently across databases, convert an entitlement into a dispute the claimant must prosecute from below. Field research in Jharkhand by Jean Dreze, Reetika Khera and their co authors, published in the Economic and Political Weekly in 2017 under the title “Aadhaar and Food Security in Jharkhand”, found the public distribution system in the study districts excluding genuine households through precisely these mechanisms, households that could not draw rations because authentication failed or seeding was incomplete, and concluded that the reform had imposed pain without a commensurate gain in the leakage it was meant to cure. The State of Aadhaar Report 2019, a national survey conducted by Dalberg and commissioned by the Omidyar Network, found both that Aadhaar had improved residents’ experience of welfare delivery in the main and that a measurable share of beneficiaries had faced denial of a benefit or service connected with an Aadhaar related failure, so that the national picture contained both effects at once. Writers such as the economist Reetika Khera have also contested the savings arithmetic itself, arguing that figures attributed to Aadhaar conflate deletions that arose from ordinary list cleaning and other reforms with savings caused by biometric identification, and so overstate what the technology contributed. Critics add the structural objection that a central authentication system, whatever the Act says about purpose limitation, creates the record of a citizen’s transactions with the State and the technical capacity for a profile, and that a benefit identity a person cannot practically decline is compulsory in substance even where the text calls enrolment an entitlement. It is part of the record that notes prepared by journalists and field researchers documenting exclusion cases in the welfare system depend on the testimony of affected families who have reason to fear official displeasure, which is why reporting on this subject makes heavy use of protected source handling of the kind described in this account of how journalists use secure workspaces to protect sources.

The two cases meet at two points where the statute itself takes a position. The first is the proviso. Parliament did not leave the exclusion objection to administrative goodwill: it wrote into Section 7 that a person to whom no number has been assigned shall be offered alternate and viable means of identification for delivery of the benefit. The delivery architecture of the large Fund financed programmes is itself organised around seeding and authentication at the point of sale or payment, as the implementation record of the Food Security Act and the public distribution system shows in detail, and the same design choice runs through the wage payments of the rural employment guarantee as implemented on the ground, where a worker’s wages travel through an identified bank or post office account and a failed match becomes a delayed wage. The proviso converts the question in any individual case from whether Aadhaar can be demanded to whether the alternative was offered, and the persistence of denial reports after the proviso suggests a gap between the statutory answer and its observance that is a failure of administration rather than of text. The second meeting point is the majority judgment, which recorded the evidence of exclusion and held that the denial of a benefit in some cases, on failure of authentication, did not make Section 7 liable to be struck down: the constitutional judgment was that conditionality with exception machinery is permissible, while the practical judgment, which the reports keep supplying, is that the machinery is where the statute succeeds or fails. A reader weighing the two cases should keep those two judgments separate, because much public argument on this Act consists of trading a constitutional point for an administrative one without noticing the exchange.

One statute among several: the data protection setting

The Aadhaar Act is a sectoral data protection code as well as an identity law, and the two descriptions interact. Its Chapter VI protects one database with rules of remarkable strictness, the absolute sharing ban on core biometrics, the purpose limits on requesting entities, the prohibition on the Authority collecting the purpose of an authentication, and criminal penalties behind all of them. Section 30 even borrows the general law, deeming biometric information to be sensitive personal data under the Information Technology Act, 2000. What the Aadhaar Act does not do is regulate personal data generally: the employer’s file, the hospital’s records and the platform’s behavioural profile lie outside it, in a gap that the Supreme Court’s privacy judgment identified and that the majority in the Aadhaar case pressed the government to fill with a data protection statute. Parliament eventually enacted a general code in the Digital Personal Data Protection Act, 2023, whose scheme, exemptions and enforcement machinery are analysed in the article on that Act. The relationship between the two statutes is complementary rather than competitive: the Aadhaar Act governs one repository and one identifier with specific rules, the general Act governs processing of digital personal data across sectors, and a provision of the Aadhaar framework, such as the Section 29 sharing ban, continues to govern its own subject with the strictness its text commands. For the researcher, the instructive point is the sequence. India built its most demanding data protection rules into a welfare delivery statute in 2016, seven years before it legislated for personal data at large, and it built them because the database came first and needed defending before a general philosophy of data protection had been agreed.

The definitional machinery: who counts, and what counts

The Act’s outcomes are decided less often by its operative sections than by its definitions, and Section 2 repays close reading because three of its choices structure everything downstream.

The first choice is the unit of entitlement. The Act does not distribute its number to citizens. It distributes it to residents, defined by presence: an individual who has resided in India for one hundred and eighty two days or more in the twelve months immediately preceding the application. The choice is functional. A delivery statute must identify the population it delivers to, which includes the long term foreign resident, the returning emigrant and the person whose citizenship papers are themselves in dispute, and it must do so without adjudicating nationality, a question the Indian state reserves to the Citizenship Act, 1955 and the constitutional provisions that precede it. Section 9 is the mirror of the same choice seen from the other side: the number or its authentication shall not, by itself, confer any right of, or be proof of, citizenship or domicile. The two provisions together mark the number’s legal character with precision. It is an identity credential for residents, generated from the fact of residence and demography, and it is evidentiary of identity only.

The second choice is what the system records. Demographic information is confined to name, date of birth, address and other relevant information specified by regulations, and is then cut down by an exclusion list of unusual specificity: it shall not include race, religion, caste, tribe, ethnicity, language, records of entitlement, income or medical history. The excluded list is a set of decisions about the state’s own knowledge. A repository that does not hold religion or caste cannot sort by religion or caste. A repository that holds no records of entitlement does not itself know what a person receives, which is the definitional companion of Section 32(3): the Act builds the wall between identity and entitlement twice, once by excluding entitlement data from the definition of what is collected at enrolment, and once by forbidding the Authority to keep the purpose of an authentication. Biometric information is then divided in two by the definition of its core: fingerprint, iris scan and any other biological attribute specified by regulations are core biometric information, while the photograph, which is also biometric information within the wider definition, sits outside the core and so outside the absolute sharing ban of Section 29(1), a calibration that lets the photograph travel for verification display while fingerprints and iris scans may not travel at all.

The third choice is the breadth of the delivery vocabulary. “Subsidy” covers any form of aid, support, grant, subvention or appropriation in cash or kind. “Benefit” covers any advantage, gift, reward, relief or payment in cash or kind. “Service” covers any provision, facility, utility or other assistance in any form. Each definition ends with a power in the Central Government to notify further items into the class. Read alone, the three definitions would support conditionality across nearly the whole field of the state’s dealings with the individual. Read with Section 7, they do not, because the funding condition qualifies the composite phrase and confines the operative power to Fund financed delivery. The definitional breadth and the funding narrowness are both deliberate, and the tension between them is the statute’s central interpretive fact: whenever a demand is challenged, the argument is almost never about whether the thing demanded is a benefit or service, and almost always about whether its expenditure is incurred from the Consolidated Fund. Researchers preparing the Act for examination or litigation should expect the decided cases to arrange themselves along that seam.

One more definition completes the set. “Records of entitlement” means records of benefits, subsidies or services provided to or availed by an individual under any programme, and the Act places such records where they belong, with the programme authorities that generate them in the course of delivery, expressly outside the demographic information the repository holds. The scheme databases know what a household drew; the identity repository knows who the person is; authentication is the controlled meeting of the two, in which the repository confirms a match and learns nothing about the entitlement, and the scheme learns that identity is established and receives no biometrics. That division of knowledge, visible only in the definitions and in Sections 29 and 32, is the Act’s answer in advance to the charge that it builds a single database of the citizen’s dealings with the state.

How conditionality is switched on in practice

Section 7 confers a power; it does not exercise it. The power is exercised scheme by scheme, by notification, and the practical law of Aadhaar conditionality is therefore a body of notifications issued under the section by the Central Government and the State Governments for individual programmes. In the first months of 2017 the Union ministries issued a large batch of such notifications across the welfare field, bringing the principal centrally sponsored schemes within the section and specifying dates by which beneficiaries were to enrol or seed their numbers, and States issued their own notifications for State schemes. Each notification is the operative legal instrument for its scheme, and the first question in any dispute about a demand is a documentary one: is there a notification under Section 7 for this scheme, and what exactly does it require, authentication, proof of possession, or an application for enrolment.

Behind the notification sits a chain of operations that the statute authorises but does not itself describe, because Section 23 leaves the processes to the Authority’s regulations. The beneficiary’s Aadhaar number is first seeded into the scheme’s own database, matched against the name and account details the scheme already holds, a step at which spelling differences between the scheme record and the Aadhaar record generate a class of failures that has nothing to do with biometrics. Wages and benefits are then paid through identified accounts. In the public distribution system, the household’s entitlement is drawn at the fair price shop after the beneficiary authenticates at the shop’s point of sale device, or establishes identity by an accepted alternative where the framework provides one. In the cooking gas subsidy, the subsidy amount is transferred to the consumer’s bank account after purchase at the market price, the identification work having been done by seeding rather than by repeated authentication. In the rural employment guarantee, the worker’s job card and bank or post office account are seeded with the number, wages are credited through the payment system on the strength of that seeding, and the rural development administration moved the programme’s payments onto the Aadhaar based payment system, in which the number itself functions as the financial address that routes the wage to the account mapped against it in the payments infrastructure. The three schemes thus use the same statute in three different intensities: authentication at every transaction, seeding once with transfer thereafter, and seeding as the routing key for payment. The exclusion risks differ accordingly, concentrating at the point of repeated authentication, which is why the ration shop, and not the gas agency, generated the documented denial cases.

The statutory answer inside this machinery is the proviso, and the machinery’s designers built exception handling around it in the delivery systems: alternative documents for the unnumbered, exception registers at delivery points, and manual overrides recorded by the delivering official. Whether those mechanisms function at a given shop on a given day is an administrative fact that varies, and the variance is the substance of the exclusion findings reported in the contested evidence section of this article. What a reader should carry away is the distribution of legal responsibility the design creates. The Act supplies the power and the proviso. The notification activates the power for a scheme. The regulations and the scheme guidelines supply the exception machinery. A denial that occurs because a notification was never issued is a defect of the second layer. A denial that occurs because the proviso was not honoured is a defect of the last layer, and it is a breach of the statute itself rather than a hardship the statute happens to have caused.

The ecosystem around the repository

The Act names a small cast of actors and lets regulations multiply their roles. Registrars are entities authorised or recognised by the Authority for enrolling individuals, in practice chiefly State Governments and Union territory administrations, public sector banks, the postal department and other public bodies, which in turn engage enrolling agencies to do the capturing. The requesting entity is the agency or person that submits a number with demographic or biometric information to the repository for authentication. Around the requesting entity, the authentication regulations constructed the operating tiers through which authentication traffic actually flows: requesting entities connect as authentication user agencies, and reach the repository through authentication service agencies that provide the secure connectivity, while a parallel track provides an electronic know your customer service under which, on the holder’s consent and a successful authentication, specified demographic information and the photograph are returned electronically to the entity, sparing the holder the paper trail that the paper system demanded. After the judgment and the 2019 amendment, the cast acquired a new member, the offline verification seeking entity, the agency that verifies identity from the holder’s digitally signed record or code without touching the repository, and Chapter VIA gave the entire ensemble a collective name, the Aadhaar ecosystem, and made every member of it liable to the civil penalty regime for non compliance with the Act, the rules, the regulations or the Authority’s directions under Section 23A.

The grievance side of the ecosystem deserves a sentence of its own, because Section 23 lists among the Authority’s functions the setting up of facilitation centres and a grievance redressal mechanism for individuals, Registrars, enrolling agencies and other service providers. The holder who believes a record is wrong has the Section 31 alteration route, the holder who wants to see what the repository holds has the access rights of Sections 28 and 32, the holder wronged by an offence has, after the 2019 amendment, a personal right of complaint under Section 47, and the holder failed by a delivery system has the scheme’s own grievance machinery backstopped by the proviso. The Act does not create an ombudsman or a compensation code for wrongful denial, and that absence, rather than any provision actually in the text, is the most common legitimate criticism of its remedial design.

The number itself: randomness, masking and the virtual identity

Section 4’s two properties of the number, that it is random and that it shall not be re assigned, do quiet work. A random twelve digit number, generated without reference to the holder’s attributes or identity, leaks nothing by being overheard: it encodes no birth date, no region of enrolment, no community marker, and no sequence from which neighbours in the register can be inferred. The twelfth digit functions as a check digit, computed by an error detecting algorithm so that a mistyped or misheard number is caught at entry rather than matched against the wrong record. Non re assignment closes the register’s history: a number issued to a person remains that person’s across the register’s life, and is deactivated rather than recycled when it leaves use, so that an old reference in any scheme database can never resolve to a different human being.

Two devices built under the regulatory framework extend the same logic of minimum exposure. The masked Aadhaar, a downloadable form of the letter, displays only the last four digits of the number, the rest being obscured, so that the copy a person hands to a hotel or an office proves the fact of the credential while withholding the number itself. The virtual identity is a temporary, revocable sixteen digit random number mapped to the holder’s Aadhaar number, which the holder may generate and use in place of the number in authentication and verification transactions, so that the identifier exposed to the transacting entity is one the holder can withdraw and regenerate. Both devices answer the same design problem the statute leaves to its administrators: the number must be usable everywhere it is legitimately required, and yet its mere possession by every transacting office would build, across the offices, a joinable trail, so the framework supplies ways of proving identity that minimise what the proof discloses.

Enrolment across a lifetime: children, updates and failing fingerprints

A biometric register of a whole population must solve a problem a document register avoids: the evidence it relies on changes with the body. The Act’s instrument is Section 6, which empowers the Authority to require holders to update demographic and biometric information from time to time to ensure continued accuracy. The enrolment framework builds the life cycle on it. Children below five years are enrolled on the basis of demographic information and a photograph, without biometric capture, the record being linked to a parent’s number, and the document issued for such a child is the blue coloured Baal Aadhaar. The child’s biometrics are captured when the child attains five years, and a further mandatory biometric update falls due at fifteen years, because the fingerprints and iris patterns of a young child are not stable enough to serve as a lifetime key. Holders who miss the update windows face deactivation of the child’s number until the update is completed, a design that trades a measure of compulsion for the register’s continuing reliability.

At the other end of the life cycle, and in the manual workforce of every age, sits the worn fingerprint. Manual labour abrades the ridges on which fingerprint matching depends, age flattens them, and injury and skin conditions interfere further. The enrolment framework’s exception process, which records a biometric exception for the person who cannot give a complete set, has its delivery counterpart in the recognition that authentication for such a person must be capable of resting on another factor: the iris where the fingerprint fails, demographic or one time password based verification where the framework permits it, and the exception and override procedures of the delivery scheme where none of those succeed. Section 6 updates serve the same population in a longer frame: a holder whose biometrics have degraded can re capture them at an enrolment centre, refreshing the template against which future authentications are judged. The statute distributes the burden of bodily change between the system’s machinery and the citizen’s duty to keep the record current, and the fairness of that distribution in any particular case is precisely the ground on which exclusion disputes are fought.

What the Act does not do

A pillar article owes its reader the negative space, because misdescription of this Act generally consists of crediting it with provisions it lacks.

It does not make enrolment compulsory. Enrolment under Section 3 is framed as an entitlement of the resident, and no offence, penalty or disability attaches under this Act to the mere fact of not holding a number. Compulsion enters only through the side doors the judgment mapped: a Section 7 notification for a Fund financed scheme, a specific provision of another statute such as Section 139AA for the permanent account number, or, in the period before the judgment, administrative instruments that the Court then struck down or confined. An article or an official who says that the Aadhaar Act makes Aadhaar compulsory, without saying for what and by which instrument, has not described the Act.

It does not make the number a proof of citizenship, domicile or entitlement. Section 9 excludes the first two expressly, and the third is excluded by the structure: possession of a number establishes identity within the system, and whether a person qualifies for a scheme is a question for the scheme’s own eligibility rules, decided by the scheme authority on the scheme record. It does not authorise the Authority to profile: Section 32(3) forbids it to collect, keep or maintain information about the purpose of authentication, and Section 29 forbids it to share the biometric core at all. It does not regulate the databases of others: State resident data hubs, scheme registers and private databases are governed by their own instruments, and the Act’s protections bind the repository and the ecosystem the Act creates, not every database in which a number happens to be stored, a boundary the litigation about convergence of databases brought into view and which the general data protection statute enacted later polices in its own terms.

It does not provide compensation, an appeal tribunal or a limitation free remedy for the person wrongly denied. The wrongly denied beneficiary’s remedies run through the scheme’s grievance machinery, the proviso’s enforceable promise of alternative identification, the Authority’s alteration and access routes for record errors, and the constitutional courts. Parliament built a delivery statute with a criminal code for misuse and, after 2019, a civil penalty code for ecosystem non compliance, but it did not build a damages remedy for exclusion, and the debate about whether that gap should be filled is a live legislative question the text leaves open.

The distinctions that decide questions

Most errors about this Act dissolve into one of five distinctions, and an aspirant or researcher who keeps them separate can answer nearly every question the statute generates.

Enrolment is not compulsion. Getting a number is an entitlement exercisable by any resident; being required to use it is a separate event that must be traced to Section 7, to another statute, or to nothing at all, in which case the demand is unlawful. Authentication is not offline verification. Authentication is a live transaction with the repository under Section 8, consent based and answered without releasing core biometrics; offline verification under Section 8A never touches the repository and carries its own prohibition on the verifying entity collecting or storing the number or biometrics. Identity is not entitlement. The repository certifies a match; schemes decide eligibility; Section 9 adds that the match proves neither citizenship nor domicile. The identity information is not the biometric core. Demographic information may move, under regulation and consent, through the channels the Act permits; core biometric information may not be shared with anyone for any reason whatsoever, a distinction Section 29 draws in consecutive sub sections and the courts have treated as the Act’s signature safeguard. Finally, the Authority’s complaint gate is not closed. Before the judgment, Section 47 confined cognizance of offences to complaints by the Authority; after the judgment and the 2019 substitution, the individual’s complaint stands beside the Authority’s, and a penalty map of the Act that still shows the Authority as sole complainant is a map of a statute that no longer exists.

The interim years: the programme under court orders before the judgment

Between the first challenge and the final judgment, the programme lived under a sequence of interim orders that functioned, for six years, as a provisional constitution for the scheme, and a reader of the Act should know them because several positions commonly attributed to the Act itself are in fact the residue of those orders.

The first petitions were filed in 2012, among them the writ petition of Justice K.S. Puttaswamy, a retired judge of the Karnataka High Court, whose name the litigation carries. In September 2013 the Supreme Court directed, in interim terms, that no person should suffer for not getting an Aadhaar card in spite of a circular or rule making it mandatory, and that the government should ensure that the number was not issued to illegal immigrants: the twin poles of the whole later dispute, compulsion and inclusion, were set in the first order. In March 2014 the Court restrained the Authority from sharing resident information with any agency without the resident’s written consent, responding to an episode in which data had been sought for a criminal investigation. In August 2015 a three judge bench referred the privacy question to a Constitution Bench and, in the meantime, confined the use of the number to the public distribution system and the cooking gas subsidy, and then only voluntarily. In October 2015 a five judge bench widened the voluntary field, at the government’s request, to include the rural employment guarantee, pension schemes, the employees’ provident fund and the financial inclusion programme, while repeating that enrolment and use must remain voluntary and must not be insisted upon until the Court decided. The Union then took the legislative route, and enacted the 2016 Act while those orders were in force, a sequence the petitioners criticised as an attempt to outflank the Court and the government defended as its legislative right. The final judgment recorded, at headline level, that the Act did not violate the interim orders, and the orders dissolved into the judgment’s own holdings. Their importance in the pillar’s account is that the settlement they improvised, voluntary enrolment, conditionality confined to funded welfare, no denial for want of the number, is recognisably the settlement the judgment and the 2019 amendment eventually made permanent.

Section 57: the rise and fall of the private use clause

For three years, the most commercially important words in the Act were its third last section. Section 57, as enacted, provided: “Nothing contained in this Act shall prevent the use of Aadhaar number for establishing the identity of an individual for any purpose, whether by the State or any body corporate or person, pursuant to any law, for the time being in force, or any contract to this effect: Provided that the use of Aadhaar number under this section shall be subject to the procedure and obligations under section 8 and Chapter VI.” The drafting form is that of a savings clause, a provision that merely removes a possible objection. Its commercial effect was that of an enabling Act. Read with the authentication machinery and the electronic know your customer service, it allowed any company whose contract with a customer said so to build Aadhaar authentication into onboarding: the telecom operator verifying a subscriber, the bank opening an account, the financial technology firm enrolling a borrower, the employer checking a worker. An industry of private authentication grew up under a clause that its defenders described as permissive and its critics as the mechanism by which an identity built for welfare delivery was converted into general purpose infrastructure for the private economy, without Parliament ever debating a private use statute.

The Court’s treatment of the clause was the judgment’s most consequential commercial holding. The majority struck down the part of Section 57 that permitted use by any body corporate or person pursuant to a contract, holding it unconstitutional: a provision that let private parties make a biometric identity check the price of a private transaction lacked the safeguards and the legislative deliberation that such a power required, and, on the majority’s analysis, enabled the profiling the Act elsewhere forbade. The strike did not depend on the clause’s abuse in any particular case. It rested on the clause’s open texture, its “any purpose” and “any contract”, which placed no legislative limit on the ends for which a body corporate could demand authentication. Justice Bhushan’s concurring conclusions struck the section’s last phrase, “or any contract to this effect”, to the like effect. Between September 2018 and July 2019 the private authentication economy accordingly lost its legal basis, and entities that had built on it reverted to documentary verification. The 2019 Act then did two things at once: it omitted Section 57, completing the judgment’s work on the text, and it built the gated replacement described in the amendment section, under which an entity may authenticate only if it meets the Authority’s privacy and security standards and is permitted by a parliamentary law or by a prescription in the interest of the State. The sequence is the cleanest case in the Act’s history of the Court and Parliament dividing the labour of legislation: the judgment identified the constitutional defect, open endedness, and the amendment supplied the limiting principle, consent plus a legislative or expressly prescribed gateway, that the original clause had omitted.

The disclosure gateways in practice

Section 33 deserves a second, slower pass, because it is the place where the Act’s privacy code meets the ordinary demands of the state for investigative access, and because its design contains the only executive controlled exception to the confidentiality scheme.

The court gateway as enacted allowed disclosure pursuant to an order of a court not inferior to that of a District Judge, with a hearing to the Authority before the order is made. Its weakness, which the litigation exposed, is that the person whose identity information or authentication records are sought had no statutory right to be heard at all: the Authority was heard, the applicant for disclosure was heard, and the subject of the record was not a party to the conversation. The majority read the provision down to build the subject in, holding that the individual must be given an opportunity of hearing before disclosure is ordered. Parliament completed the repair in 2019 by raising the forum to a Judge of a High Court, extending the hearing to the concerned Aadhaar number holder in terms, and adding the proviso that core biometric information shall not be disclosed under the sub section at all. That last addition interacts with Section 29 in an instructive way: the court gateway never in fact overrode the core biometric ban, since Section 33(1) suspends only Section 28(2) and (5) and Section 29(2), but the 2019 proviso put the point beyond construction by stating it in the gateway itself.

The national security gateway has a different design and drew a different fate. As enacted, a direction by an officer not below Joint Secretary, specially authorised, could require disclosure in the interest of national security, subject to prior review by an Oversight Committee composed of the Cabinet Secretary and the Secretaries of the Departments of Legal Affairs and of Electronics and Information Technology, with each direction valid for three months and renewable once after review. The majority struck the provision down: a mechanism in which the executive reviews the executive, without any judicial member or any parliamentary anchor, did not supply the safeguard a national security exception to a privacy code requires, and the provision was too widely framed for the interest it protected. The 2019 amendment then re enacted the gateway in hardened form, raising the directing officer to the rank of Secretary to the Government of India, which is the form the provision takes in the statute book as amended, with the Oversight Committee and the three month validity retained. The two gateways thus stand in instructive contrast: the judicial one was read down and strengthened, the executive one was struck down and rebuilt at a higher rank, and the core biometric information remains outside both.

Aadhaar in the tax law: the Section 139AA bridge

The one compulsory use of Aadhaar outside Section 7 that survived scrutiny does not live in the Aadhaar Act at all. It lives in the Income tax Act, 1961, into which the Finance Act, 2017 inserted Section 139AA with effect from 1 July 2017. The section required every person eligible to obtain an Aadhaar number to quote it in the application for allotment of a permanent account number and in the return of income, required existing permanent account number holders who were eligible to obtain Aadhaar to intimate the number to the prescribed authority by the notified date, and provided that, on failure, the permanent account number would be deemed to have been not issued. The provision was challenged before the Aadhaar Act challenge was decided, and in Binoy Viswam v Union of India in 2017 the Supreme Court upheld the section under the taxing power while reading down its consequence: the invalidity of a permanent account number for failure to link would operate prospectively and the section would not be applied to persons not otherwise required to enrol. The Aadhaar bench of 2018 then examined the provision again against the privacy standard and upheld it, placing the permanent account number linkage with the Fund financed welfare uses as the two standing instances of lawful compulsion.

The bridge matters to the pillar’s thesis because it proves the thesis from the other side. Section 139AA is a direct parliamentary command in a revenue statute, justified by the documented abuse of multiple and false permanent account numbers in tax evasion. It did not need Section 7, because it was not a welfare delivery provision, and it did not need Section 57, because it was itself the law pursuant to which the number was used. Where a compulsory demand rests on a parliamentary text that states the duty and the consequence, the constitutional analysis proceeds on proportionality grounds about the duty itself. Where it rests on a circular, a rule, or a contract clause, it fails at the earlier question of authority. That is the whole law of Aadhaar conditionality in one comparison, and it is why the artifact table’s source column, rather than its yes and no column, is the column to study.

One fact, two consequences: the Fund as the Act’s organising idea

The series to which this article belongs has repeatedly found a statute’s reach determined by a hook inside a definition or a funding clause rather than by the grandeur of its subject, and this Act is the pattern’s purest instance. The Aadhaar Act presents itself as a law about identity, the most universal of subjects, since everyone has one. Its legal footprint is fixed by a concept from public finance. Because the conditionality clause is keyed to expenditure from the Consolidated Fund, the Act can compel identification at the ration shop and at the wage payment, and cannot compel it at the school gate, the examination hall, the telephone shop or the bank counter. Because the same funding connection supplied the incidental matter that made the Bill a Money Bill, the definition of the Act’s reach also supplied the constitutionally contested definition of its own enactment procedure, and the two questions that dominate its history, “for what may Aadhaar be demanded” and “was the Act validly passed”, are at bottom one question asked in two courts, the delivery point and the Constitution Bench. A reader who grasps that the Consolidated Fund is the organising idea of the statute, and not a piece of accounting decoration inside Section 7, has the key that the Act’s structure, the judgment’s pattern of holdings and the dissent’s objection all turn on, and has it in the form the Act itself supplies.

The federal frame: two governments inside one section

Section 7 is drafted in the alternative throughout: “The Central Government or, as the case may be, the State Government may… require”. The formulation reflects the delivery reality that India’s welfare state is administered jointly. The Union designs and part finances the great national programmes, the States administer them through their own machinery, and the States run extensive schemes of their own, pensions, scholarships, housing and subsidy programmes financed from State budgets. The 2016 text, however, keyed the condition to the Consolidated Fund of India alone, which left a drafting puzzle for State schemes financed from a State’s own Consolidated Fund, and the 2019 amendment solved it by inserting the State fund into Section 7 and into the long title, so that the two funding sources stand side by side in the section. States appear elsewhere in the machinery as well: they are the principal Registrars through whom enrolment is conducted, their governments are parties to the memoranda of understanding under which enrolment and authentication proceed, and State notifications under Section 7 are the operative instruments for State schemes. The design is therefore neither a Union imposition on State delivery nor a State free for all: Parliament wrote the power, the proviso and the protection code at the centre, and the two levels of government exercise the power within their own spheres, which is why a reader researching a particular demand must always ask which government issued the notification and from which fund the scheme is paid. The Money Bill controversy has a federal shadow here too, since a Bill routed past the Council of States is, by design, a Bill on which the States’ House could not insist, and the charge that the route diminished the federal chamber is the institutional form of the same objection the petitioners pressed in court.

The proviso tested: denial, exception registers and what the record shows

The practical meaning of the proviso is best understood through the cases that made it a public question. The best documented pattern is the cancelled or unseeded ration card. Where the seeding drive linked ration cards to the numbers of household members, cards that remained unlinked, because a member had no number, because the seeding failed, or because the household could not complete the process, were cancelled or suspended in several States, and households found their entitlement extinguished not at the authentication step but before it, at the database step. The case that fixed the pattern in the national record occurred in Simdega district in Jharkhand in September 2017: an eleven year old girl died in a household whose ration card had been cancelled for want of Aadhaar linkage, her family and activists attributing the death to starvation after months without rations, and the district administration attributing the death to illness and disputing the connection. Whatever the resolution of that dispute in the individual case, the cancellation itself was not disputed, and the case established the proposition that matters for the statute: conditionality operates through databases before it operates through fingerprints, and a database error is not answered by the proviso unless the alternative identification route is actually restored to the household. The Union Government’s responses to the denial reports of that period ran in the proviso’s own language: directions that no beneficiary be denied for want of Aadhaar, that exception registers be maintained at delivery points, and that alternative documents be accepted where authentication failed. The delivery framework’s exception machinery, exception registers, alternative documents, manual override by the delivering official, is thus not an embellishment of the statute but the administrative form of its proviso, and its uneven functioning is the gap between the Act as written and the Act as experienced that the evidence sections of this article have kept in view.

Two enforcement codes: prosecution and civil penalty

The Act after 2019 enforces itself through two distinct codes, and the distinction maps the two kinds of wrong it anticipates. The criminal code of Chapter VII addresses conduct: impersonation, unauthorised collection, disclosure, unauthorised access to the repository, tampering, misuse by a requesting entity. Its sanctions are imprisonment and fines, its investigation is by police officers of Inspector rank and above, its forum is the court of a Chief Metropolitan Magistrate or Chief Judicial Magistrate or above, and its trigger, as amended, is a complaint by the Authority or by the individual. The civil code of Chapter VIA addresses compliance: an entity in the Aadhaar ecosystem that fails to comply with the Act, the rules, the regulations or the Authority’s directions, or fails to furnish information the Authority requires, faces money penalties assessed on a regulatory scale, up to one crore rupees per contravention plus a daily rate for continuance, adjudicated inside the regulatory system by an Adjudicating Officer with an appeal to a specialised tribunal, and recoverable as arrears of land revenue if unpaid. The philosophy is the standard one of modern Indian regulatory statutes: crime for the intruder and the dishonest insider, penalties for the licensed participant who cuts corners. A reader assessing an Aadhaar controversy should therefore begin by classifying it, because a data leak by an ecosystem entity is ordinarily a Chapter VIA matter with a compensation shaped outcome, a hacking of the repository is a Section 38 crime with a ten lakh rupee minimum fine, and a misuse of a person’s information by an insider can be both at once. The one enforcement instrument the Act still lacks is a direct compensation route for the individual harmed by exclusion, a gap recorded in the negative space section of this article and left by both the original Act and the 2019 amendment unfilled.

The chronology of the statute book

The sequence, gathered in one place, runs as follows. The Authority is created by executive notification on 28 January 2009. The first number is issued on 29 September 2010. The National Identification Authority of India Bill is introduced in December 2010 and reported against by the Standing Committee on Finance in December 2011. The first interim orders confine compulsion from September 2013 onward. The Bill is introduced in the Lok Sabha on 3 March 2016, passed there on 11 March, returned with recommendations by the Rajya Sabha on 16 March and not amended, and assented to on 25 March 2016 as Act 18 of 2016. Its provisions commence in two tranches, the Authority and miscellaneous provisions on 12 July 2016 and the remainder on 12 September 2016. The nine judge bench decides the privacy question on 24 August 2017, holding privacy a fundamental right, and the five judge bench decides the Aadhaar challenge on 26 September 2018, upholding the Act with the deletions recorded in this article. An amending Bill passes the Lok Sabha on 4 January 2019 and lapses with the House; an Ordinance carries the scheme from 2 March 2019; the Aadhaar and Other Laws (Amendment) Act, 2019, Act 14 of 2019, is assented to on 23 July 2019 and brought into force on 25 July 2019. The Rojer Mathew bench refers the Money Bill questions to a larger bench on 13 November 2019, and the review petitions against the Aadhaar judgment are dismissed by a majority of four to one in January 2021. The statute a reader consults after those events is a palimpsest: a 2016 text, minus the provisions the Court removed, plus the consent architecture the 2019 Parliament built, with one constitutional question, the Money Bill route by which the whole structure was enacted, still awaiting the seven judge bench to which it stands referred.

The holder’s own rights in the record

Against the repository’s enclosure, the Act gives the holder a defined bundle of rights in his or her own record, and they are worth stating together because they are scattered across chapters.

The right to be told comes first in time. At enrolment, Section 3(2) obliges the enrolling agency to inform the individual of the manner in which the information will be used, of the nature of recipients with whom it is intended to be shared during authentication, and of the existence of a right to access the information, the procedure for making requests, and the officer or department in charge to whom requests may be made. At every authentication, Section 8(3) renews the disclosure in transactional form: the requesting entity must state the nature of information that may be shared upon authentication, the uses to which the received information may be put, and the alternatives to submitting the information at all. The right of access follows. Under the proviso to Section 28(5), a holder may request the Authority to provide access to his or her identity information, excluding core biometric information, in the manner specified by regulations. Under Section 32(2), every holder is entitled to obtain his or her authentication record, the log of when authentication was sought, by which requesting entity, and with what response, so that the one person with the strongest interest in detecting a misuse is equipped to detect it. The right of correction stands beside access: Section 31 entitles the holder to request alteration of demographic information that is incorrect or has changed, and of biometric information that is lost or has changed, and obliges the Authority, if satisfied, to make the alteration and intimate it to the holder, while forbidding any alteration of identity information in the repository except in the manner the Act or regulations provide, a restriction that protects the holder against silent third party changes as much as it binds the holder’s own requests.

The retention of the authentication record is the point at which the holder’s access right and the system’s memory meet, and the judgment fixed its limit. Section 32(1) leaves the period for which authentication records are maintained to the regulations, and the authentication regulations as first framed permitted the records to be archived for five years. The Supreme Court struck that provision down, holding that retention of authentication records for a period longer than six months is not permissible, so that the log the holder may inspect under Section 32(2) is a record of the recent past rather than a lifetime ledger of transactions. The limit cuts both ways, and both ways are instructive. It confines the state’s ability to reconstruct a citizen’s dealings from the repository’s side of the wire, which is the surveillance concern the Act’s design addresses at every level. It also confines the holder’s own audit trail, so that a misuse discovered late may have passed beyond the retained record, a cost the Court evidently regarded as the necessary price of the first benefit. Read with Section 32(3), which forbids the Authority to keep the purpose of any authentication, the retention ceiling completes a consistent scheme: the repository remembers a transaction briefly, records nothing of its reason, and discloses the fact of it, within the retained period, to the one person the fact concerns.

Two further rights are remedial. The holder who is the victim of an offence under the Act may, after the 2019 substitution of Section 47, file the complaint on which a court takes cognizance, in place of waiting on the Authority to move. And the holder enrolled as a child has the right, created by the 2019 amendment, to apply within six months of attaining eighteen years for cancellation of the number issued in childhood, the only point in the Act at which the subject of the record is given a unilateral exit from it. What the bundle does not contain should be recorded with the same care: no right to require the Authority to delete the record at will in adult life, no right to be told the purpose for which another party sought authentication beyond the record of the fact, and no statutory compensation for the misuse the bundle is designed to reveal. The rights are transparency and correction rights, well adapted to a register whose chief risks are error and unauthorised use, and silent, as this article has noted, on the distinct risk of exclusion by the delivery systems that consume the register’s answers.

Testing a demand for Aadhaar: a worked method

The statute equips a reader to test any concrete demand for Aadhaar by four questions, in a fixed order, and the method is worth setting out because it converts the article’s argument into a usable instrument.

The first question is the instrument. What legal document creates the demand? If the answer is a notification under Section 7, the demand is one the Aadhaar Act itself authorises, subject to the remaining questions. If the answer is a provision of another statute, such as Section 139AA of the Income tax Act, the demand stands or falls by that statute. If the answer is a departmental circular, a condition in a form, an instruction from a head office or a clause in a private contract, the demand rests on the class of instruments the judgment disabled: the mobile linking circular was set aside, the money laundering rule for bank accounts failed, and the contract route in Section 57 was struck down and the section omitted. A demand whose instrument cannot be produced cannot be improved by insistence.

The second question is the funding, and it arises only for demands said to rest on Section 7. Is the subsidy, benefit or service one for which the expenditure is incurred from, or the receipt forms part of, the Consolidated Fund of India or, after the 2019 amendment, of the State? The answer is a fact about the scheme’s financing, not about its benevolence or its administering ministry. A scholarship, a pension, a ration and a wage under the employment guarantee answer yes. A school place, a university examination and a seat in a competitive test answer no, and with the funding answer the Section 7 analysis ends.

The third question is the alternative. Where the demand is lawful, what must the demander offer the person who has no number, or whose number cannot be authenticated? Under Section 7 the answer is written into the proviso: alternate and viable means of identification for delivery of the benefit, and for children, after 2019, the stronger rule that the child shall not be denied for failure to authenticate or to produce. In the voluntary sectors reopened in 2019, the answer is the parallel set of officially valid documents and modes which the Telegraph Act and the money laundering framework require to be made available, with the rule that no person be denied the service for not having a number. A lawful demand is therefore always a demand with an exit, and a demand presented without one misstates its own legal basis.

The fourth question is the remedy, and it depends on which side of the counter the reader stands. The person wrongly denied invokes the proviso through the scheme’s grievance machinery and, failing that, the constitutional courts, which have consistently treated the right to receive a subsistence benefit as an aspect of the right to life. The person whose information is misused invokes the correction and access rights, complains under Section 47, and sets the Authority’s civil penalty jurisdiction in motion against an erring entity of the ecosystem. The provider or employer who receives a demand from above applies the same four questions upward, and the statute protects the refusal that the questions justify. This article’s purpose is discharged if its reader, meeting the next demand for Aadhaar in whatever capacity, asks first for the instrument, and does not stop asking until the funding, the alternative and the remedy have each been produced.

Who pays for authentication

Section 8(1) provides that the Authority shall perform authentication subject to such conditions and on payment of such fees, and in such manner, as may be specified by regulations, and Section 23 includes among the Authority’s functions the levying and collection of fees, or the authorisation of Registrars, enrolling agencies and service providers to collect them. The authentication query is therefore a priced service, and the requesting entity, not the holder, is the customer: a bank, a telecom operator, a government department consumes an answer from the repository and pays the Authority for it under the fee schedule the regulations fix, with enrolment of residents remaining a programme cost borne through the Authority’s grants and revenues rather than a charge on the person enrolled. The arrangement matters to the Act’s economics in two ways. A per transaction price disciplines the demand for authentication, since an entity that pays for every query has a reason to ask only the queries it needs, and the fee stream, credited after the 2019 amendment to the Unique Identification Authority of India Fund, gives the Authority a revenue base partly independent of annual appropriations, matched against the costs of maintaining a repository that must answer queries at national scale without interruption. The holder’s side of the ledger is equally deliberate: the citizen is never the paying party to the system that identifies him or her, a choice consistent with Section 3’s framing of the number as an entitlement. Where a delivery point or an intermediary attempts to shift the cost of authentication or enrolment onto the beneficiary, the demand finds no support in the Act, and the attempt belongs to the class of informal levies that the scheme guidelines and the grievance machinery exist to correct.

Frequently Asked Questions

Q: What does the Aadhaar Act 2016 provide?

The Act gives statutory form to the Aadhaar identification system. It entitles every resident to obtain a unique twelve digit Aadhaar number on submitting demographic and biometric information, establishes the Unique Identification Authority of India to enrol residents and maintain the Central Identities Data Repository, regulates authentication, restricts the sharing and use of the information collected, with an absolute ban on sharing core biometric information, creates offences and penalties for misuse, and in Section 7 permits the Centre or a State to require authentication or proof of possession of a number as a condition for a subsidy, benefit or service financed from the Consolidated Fund.

Q: Why was the Aadhaar Act passed as a Money Bill?

The government certified the Bill as a Money Bill because its operative provision, Section 7, governs the delivery of subsidies, benefits and services whose expenditure is incurred from the Consolidated Fund of India, a subject Article 110 of the Constitution reserves to Money Bills, with the identification machinery treated as incidental to that expenditure. The certificate meant the Rajya Sabha could recommend amendments but could not insist on them, and it did recommend changes which the Lok Sabha declined. A majority of the Supreme Court later upheld the certification, a dissent called the route a fraud on the Constitution, and the wider Money Bill question stands referred to a seven judge bench.

Q: When can Aadhaar be made mandatory?

Aadhaar can be made mandatory through Section 7 of the Act, by government notification, as a condition for receiving a subsidy, benefit or service whose expenditure is incurred from, or receipt forms part of, the Consolidated Fund of India or of the State. It can also be made mandatory by a specific provision of another Act of Parliament, the standing example being Section 139AA of the Income tax Act, 1961 for the permanent account number. Outside those routes it cannot. The Supreme Court struck down compulsory linkage for bank accounts and mobile connections, and held that Aadhaar cannot be required for school admission or for examinations conducted by bodies such as the CBSE, NEET and UGC.

Q: Who can access Aadhaar authentication records?

The Authority maintains authentication records, which show the time of an authentication, the identity of the requesting entity and the response given, and the Aadhaar number holder is entitled under Section 32 to obtain his or her own authentication record. Identity information and authentication records are otherwise enclosed by Sections 28 and 29: they may be disclosed only through the gateways of Section 33, namely an order of a Judge of a High Court, after the holder and the Authority are heard, or a direction in the interest of national security by an officer not below the rank of Secretary, reviewed by the Oversight Committee. Core biometric information is never disclosed.

Q: What happened to the provision on private use of Aadhaar?

Section 57, which provided that nothing in the Act prevented the use of the Aadhaar number for establishing identity for any purpose by the State or any body corporate or person, pursuant to any law or any contract, was the provision under which private authentication had grown. The Supreme Court struck down its private use limb, holding unconstitutional the use of Aadhaar by a body corporate or person pursuant to a contract. Parliament then omitted Section 57 altogether in the 2019 amendment, and rebuilt private use on a gated, consent based framework in which an entity may authenticate only if it meets the Authority’s privacy and security standards and is permitted by a parliamentary law or prescribed in the interest of the State.

Q: What did the 2019 Aadhaar amendment change?

The Aadhaar and Other Laws (Amendment) Act, 2019 omitted Section 57, made the use of Aadhaar outside welfare delivery voluntary and consent based, introduced offline verification through the new Section 8A while barring verifying entities from collecting or storing the number or biometrics, allowed a person enrolled as a child to seek cancellation of the number within six months of turning eighteen, extended Section 7 to the Consolidated Fund of the State and protected children from denial of benefits, raised the disclosure gateways to a High Court Judge and a Secretary, created civil penalties up to one crore rupees adjudicated by an Adjudicating Officer with appeal to the Telecom Disputes Settlement and Appellate Tribunal, opened complaints under Section 47 to individuals, and established a fund for the Authority.

Q: Can a person be denied a benefit for want of Aadhaar?

Under Section 7, a person to whom no Aadhaar number has been assigned must be offered alternate and viable means of identification for delivery of the subsidy, benefit or service, so denial merely for want of a number is contrary to the section’s proviso. The Supreme Court upheld Section 7 while recording that it could not be struck down for instances of denial on authentication failure, and insisted on the exception machinery. For children, the 2019 amendment added that no child shall be denied any subsidy, benefit or service under the section for failure to establish identity by authentication, proof of possession or an application for enrolment. Denials through seeding errors and failed matches remain an administrative problem the proviso is meant to answer.

Q: What penalties does the Aadhaar Act carry?

Chapter VII creates criminal penalties scaled to the conduct. Impersonation at enrolment or by changing a holder’s information carries imprisonment up to three years or a fine up to ten thousand rupees. Unauthorised access to the Central Identities Data Repository carries imprisonment up to three years and a fine of not less than ten lakh rupees, and disclosing identity information, impersonation offences and requesting entity misuse fall in the same range, with company fines up to one lakh rupees and a general penalty of one year or twenty five thousand rupees. Since 2019, Chapter VIA adds civil penalties on entities in the Aadhaar ecosystem of up to one crore rupees per contravention and ten lakh rupees for each day of a continuing failure.

Q: Is an Aadhaar number proof of citizenship or domicile?

No. Section 9 of the Act provides that the Aadhaar number, or the authentication of it, shall not by itself confer any right of, or be proof of, citizenship or domicile in respect of the holder. The number is issued on the basis of residence, defined as one hundred and eighty two days or more in India in the twelve months before application, and it certifies identity within the system only. A foreign national lawfully resident in India may hold a number, and an Indian citizen’s status is neither created nor proved by holding one. Citizenship questions are governed by the Constitution and the Citizenship Act, 1955, not by the Aadhaar register.

Q: Who is entitled to obtain an Aadhaar number?

Every resident is entitled to obtain an Aadhaar number by submitting demographic and biometric information through enrolment, under Section 3. A resident is an individual who has resided in India for a period or periods amounting in all to one hundred and eighty two days or more in the twelve months immediately preceding the application. The Central Government may notify other categories of individuals as entitled. Section 5 requires the Authority to take special measures for women, children, senior citizens, persons with disability, unskilled and unorganised workers, nomadic tribes and persons without a permanent dwelling house. Children below five are enrolled without biometric capture, linked to a parent, with biometrics captured at five and updated at fifteen.

Q: What is offline verification, and how does it differ from authentication?

Authentication, under Section 8, is a live transaction: the number, with demographic or biometric information, is submitted to the Central Identities Data Repository, which answers whether the submission matches its record, without releasing core biometric information. Offline verification, introduced by Section 8A in the 2019 amendment, verifies identity without any transaction with the repository, using digitally signed records or codes issued by the Authority that the holder presents. The verifying entity must obtain the holder’s consent, must inform the holder of alternatives, may use what is provided only for the verification, and is prohibited from collecting, using or storing the Aadhaar number or the holder’s biometric information for any purpose.

Q: What must a requesting entity tell a person before authentication?

Section 8 requires a requesting entity to obtain the individual’s consent before collecting identity information for authentication, and to ensure the information is used only for submission to the repository. Under Section 8(3) it must inform the individual of the nature of information that may be shared upon authentication, the uses to which the information received during authentication may be put by the requesting entity, and the alternatives to submission of identity information to the requesting entity. Section 29(3) then binds the entity: identity information in its hands may not be used for any purpose other than the one specified to the individual, nor disclosed further except with the individual’s prior consent.

Q: Can a person cancel an Aadhaar number obtained as a child?

Yes, within a defined window. The 2019 amendment inserted into the enrolment provisions the rule that a child who is an Aadhaar number holder may, on attaining the age of eighteen years, make an application to the Authority for cancellation of the Aadhaar number, within six months of attaining majority. The provision answers the objection that enrolment in childhood occurs on a parent’s decision rather than the holder’s own choice: the adult is given a personal decision point on reaching majority. Outside that window the Act does not give a holder a general right to cancel or delete the number, and deactivation otherwise occurs only under the Authority’s regulation based powers, as where mandatory biometric updates for a child remain undone.

Q: Is Aadhaar required for income tax filing and permanent account number linking?

Yes, by force of a different statute. Section 139AA of the Income tax Act, 1961, inserted by the Finance Act, 2017, requires a person eligible to obtain Aadhaar to quote the number, or the enrolment application number, in the application for a permanent account number and in the return of income, and requires existing permanent account number holders to intimate the number to the prescribed authority, failing which the permanent account number is treated under the section’s scheme as liable to become inoperative. The Supreme Court upheld the provision in Binoy Viswam in 2017, reading down its retrospective consequence, and upheld it again in the Aadhaar judgment of 2018 against the privacy challenge.

Q: Can an employer make Aadhaar a condition of employment?

An employer has no statutory power to make Aadhaar a condition of hiring. The private use provision, Section 57, was struck down in its contract based limb and then omitted, so a clause in an employment contract cannot by itself authorise authentication. After the 2019 amendment, an entity may perform authentication only if it complies with the Authority’s privacy and security standards and the authentication is permitted under a law made by Parliament or prescribed by the Central Government in the interest of the State. An employer may accept an Aadhaar card voluntarily offered by a candidate as one identity document among others, and public employment programmes financed from the Consolidated Fund, such as the rural employment guarantee, stand on the separate footing of Section 7.

Q: What happens to an Aadhaar number when its holder dies or a duplicate is found?

The number is never re assigned. Section 4 provides that an Aadhaar number issued to an individual shall not be re assigned to any other individual, so the identifier cannot migrate to another person. Section 23 empowers the Authority to omit and deactivate an Aadhaar number and the information relating to it in the manner specified by regulations, which is the machinery used where a number was wrongly issued, where a duplicate enrolment is detected and the records are resolved into one, and where the register establishes a ground for deactivation such as a child’s unperformed mandatory biometric update. A deactivated number ceases to authenticate, and the holder’s route back is the Authority’s processes for correction, update and reactivation under the regulations.

Q: Who investigates and prosecutes offences under the Aadhaar Act?

Section 45 provides that, notwithstanding the Code of Criminal Procedure, a police officer not below the rank of Inspector shall investigate any offence under the Act. Prosecution begins on complaint: as enacted, Section 47 confined cognizance to a complaint by the Authority or an officer or person authorised by it, but the Supreme Court struck down that restrictive effect and the 2019 amendment substituted the section so that a complaint may also be made by an individual. Trial is before a court not inferior to a Chief Metropolitan Magistrate or a Chief Judicial Magistrate. Civil penalties under Chapter VIA follow a separate track, adjudicated by an Adjudicating Officer appointed by the Authority, with appeal to the Telecom Disputes Settlement and Appellate Tribunal.

Q: Does the Authority record why a person authenticated?

No. Section 32(3) provides that the Authority shall not, either by itself or through any entity under its control, collect, keep or maintain any information about the purpose of authentication. The Authority maintains authentication records showing the time of the transaction, the requesting entity and the response given, and the holder is entitled to obtain that record under Section 32(2). What the record cannot contain, by statutory command, is the reason the citizen transacted: the repository knows that a match was requested and answered, and is forbidden from assembling those answers into an account of the purposes for which identity was proved. The rule is one of the Act’s structural answers to the surveillance objection.

Q: When did the Aadhaar Act come into force?

The Act received the President’s assent on 25 March 2016 and was published as Act 18 of 2016. Commencement was staggered under Section 1(3). Sections 11 to 20, 22, 23 and 48 to 59, which establish the Authority, transfer the executive era programme to it and validate past action, came into force on 12 July 2016. Sections 1 to 10 and 24 to 47, covering enrolment, authentication, the Consolidated Fund conditionality in Section 7, finance, the protection code and the offences, came into force on 12 September 2016. Section 59 deemed things done under the Planning Commission notification of 28 January 2009 and the Cabinet Secretariat notification of 12 September 2015 to have been validly done under the Act.

Q: How does the Aadhaar Act protect core biometric information?

Section 29(1) imposes the Act’s strictest rule: no core biometric information, meaning fingerprint, iris scan or other specified biological attribute, collected or created under the Act, shall be shared with anyone for any reason whatsoever, or used for any purpose other than generation of Aadhaar numbers and authentication under the Act. Authentication responses exclude core biometric information by Section 8(4), publication or public display of a number or core biometric is barred by Section 29(4), and the disclosure gateways in Section 33 do not override the ban, a point Parliament made express for court ordered disclosure in 2019. Section 30 deems biometric information sensitive personal data under the Information Technology Act, 2000, and Chapter VII criminalises unauthorised access, disclosure and tampering.