Every debate about anti-terror legislation in India eventually arrives at the same question, and the TADA POTA UAPA comparison is usually conducted as if it had a one-word answer. Is the law in force harsher than the laws that came before it, or milder? Those who want a stronger statute say the Unlawful Activities (Prevention) Act is a diluted successor that dropped the powers which made the Terrorist and Disruptive Activities (Prevention) Act and the Prevention of Terrorism Act effective. Those who want the statute curbed say the UAPA is TADA and POTA combined and worse, a permanent code assembled from the harshest parts of two discredited experiments. A TADA POTA UAPA comparison conducted power by power shows that both camps are describing the same statute from different rows of it. On the row that decides whether an innocent person is convicted, the statute in force is the mildest of the three. On the rows that decide how long an unconvicted person waits in custody, it is at least as severe as either predecessor and in one respect more severe than both. The verdict of this article is therefore a split verdict, and the deciding factor can be named exactly: it is which of two distinct harms, wrongful conviction or wrongful detention, the reader treats as the graver injury. Everything that follows is the evidence for that verdict, set out power by power so the reader can reweigh it.

The comparison has a discipline that most versions of it lack. A statute is not harsh or mild in the abstract; it is harsh or mild at a specific point of operation, in what it lets the police record as evidence, in how long it lets the state hold a suspect before committing itself to a charge, in what a judge must be persuaded of before releasing that suspect, in what the prosecution is relieved of proving, and in how long the statute itself is permitted to live. Comparing the three laws at those points, rather than trading impressions of their reputations, produces a picture that neither political camp quotes whole. It also produces a practical map. A legislator asked to restore a repealed power can see what restoration would actually add. An advocate with a live case under any of the three statutes, and prosecutions under the repealed laws remain in the system decades after those laws ended, can see which procedural rules govern it. An aspirant can see why the examination version of this comparison, which usually presents a simple progression from strict to stricter, is wrong on the row that matters most.

Three statutes, three lifespans, and the survivor’s advantage

The three laws had three different relationships with time, and the difference explains more of the comparison than any single provision. The Terrorist and Disruptive Activities (Prevention) Act began in 1985 as an Act of two years’ intended life, was re-enacted in 1987 on the same temporary footing, and survived a decade only because Parliament repeatedly amended its duration clause, in 1989, in 1991 and in 1993, before declining to renew it and letting it lapse on 23 May 1995. The full passage and renewal record is set out in the account of TADA’s enactment, renewals and lapse, and it matters here for one reason: every renewal required a government to stand up in Parliament and defend the statute’s record, so the law’s continuation was a recurring political decision with a date attached. The Prevention of Terrorism Act took the opposite route to a similar end. Enacted in 2002 after a joint sitting of Parliament, it carried in Section 1(6) a fixed duration of three years from its deemed commencement on 24 October 2001, and it was repealed with effect from 21 September 2004, roughly a month before that term would have run out, by a Parliament that had changed hands; the history of POTA’s passage and repeal traces how the repeal was paired with a simultaneous amendment of the UAPA. The Unlawful Activities (Prevention) Act has never had either kind of appointment with the calendar. It was passed in 1967 as ordinary permanent legislation against unlawful associations, decades before it carried any terrorism offence at all, and the complete guide to the UAPA records how its anti-terror content arrived later, by amendment, in 2004, 2008, 2013 and 2019.

That biography is the survivor’s advantage, and it is the central historical insight of the comparison. A temporary statute must be defended afresh to be kept, and each defence invites an audit of arrests, acquittals and abuses. A permanent statute needs no defence at all to continue; changing it requires a government to spend legislative time and political capital, while leaving it alone costs nothing. The UAPA did not survive because Parliament judged it the best of the three instruments. It survived because it was the only one of the three that never had to be re-enacted, and because each time a temporary statute fell, the powers thought indispensable were carried into it rather than allowed to fall with it. The ratchet moved in one direction for a structural reason: repeal and lapse applied to the temporary laws, while amendment applied to the permanent one, and amendment almost always meant addition.

The lifespans also fix what kind of question the comparison is. TADA and POTA are closed texts; their content is frozen at the moment of lapse and repeal, subject only to the savings that kept pending prosecutions alive. The UAPA is an open text that has been rewritten four times since POTA fell, so an honest comparison must state which version of it is being compared. The comparison below takes the UAPA in its fullest form, including the 2019 amendment that added the designation of individuals, because that is the statute a proposal to revive TADA or POTA would actually be measured against. Where a power arrived late, the year of its arrival is given, since a comparison that credits the UAPA of 1967 with the powers of 2019 would repeat the commonest error in the debate.

Each statute was also born in a specific emergency, and the emergencies explain the texts. TADA was enacted in 1985 in the shadow of the Punjab insurgency and the assassination of Prime Minister Indira Gandhi in 1984, and its design, special courts near the disturbance, confessions taken by senior officers, presumptions drawn from recovery of arms in notified areas, is the design of a statute written for a countryside insurgency in which witnesses would not come forward. Its subsequent career, extension to states with no comparable insurgency and application to ordinary crime in places far from Punjab, is the standing warning about emergency drafting. POTA’s gestation was shorter and stranger: it began as an ordinance in October 2001, weeks before the attack on Parliament in December 2001 transformed its politics, and it reached the statute book only through a joint sitting in March 2002 after the Rajya Sabha refused it, the third joint sitting in the history of the republic. Its text reads as a statute drafted against a single spectacular attack and the fear of more, with an interception code and a listing schedule aimed at networks rather than at a territorial insurgency. The UAPA’s anti-terror content, by contrast, was assembled in calmer procedural forms, by amendment bills in 2004, 2008, 2013 and 2019, the 2008 reconstruction following the Mumbai attacks of November 2008 but taking the form of additions to a standing Act rather than a new emergency code. The drafting conditions left marks that the matrix preserves: the emergency statutes are the ones with the confession power and the exceptional courts, because emergencies argue for evidence and forum shortcuts, while the permanent statute accumulated the widest definitions and the designation power, because amendments passed in ordinary sessions face no sunset and invite addition without subtraction.

Permanence has a further consequence for how a statute grows, and the UAPA’s biography is its proof. A temporary Act is reconsidered as a package: renewal debates reopen the whole text, and opponents get a single vote on everything. A permanent Act changes only by amendment, and amendments in this field have a ratchet’s direction, since each is framed as a response to an attack or a gap, moved in a climate where voting against a counter-terrorism clause carries a political price that voting for it does not. Four times the UAPA has been opened, in 2004, 2008, 2013 and 2019, and each opening added offences, powers or designations; none removed any. The 2019 amendment is the purest instance of the mechanism, since it created an entirely new power, the designation of individuals, inside an Act whose terrorism chapters were then fifteen years old and had never faced the whole-Act parliamentary and judicial scrutiny that attended the birth of TADA and POTA. The survivor’s advantage of the section’s title is therefore not merely that the UAPA avoided a sunset. It is that the UAPA’s mode of growth, accretion by amendment, systematically avoids the two moments of accountability the temporary statutes could not escape, the renewal vote and the comprehensive challenge, while enjoying the same legislative majorities that enacted them.

The three-statute power matrix

The matrix below is the article in one object. Each row is a power at which the three statutes can be compared exactly, each cell records whether the power is present, absent or modified, and the final column states the position under the law in force after all amendments. The rows are then argued in prose, because a cell can record that a power exists but not what its formulation does to the person on the receiving end of it, and in this field the formulation is usually where the severity lives.

Power	TADA (1985 and 1987)	POTA (2002)	UAPA (1967, as amended)	Position under the law in force
Confession to a police officer admissible	Yes, Section 15	Yes, Section 32	No provision	No; the ordinary evidence bar governs
Maximum pre-charge custody	180 days, extendable to one year, Section 20(4)	90 days, extendable to 180 days, Section 49(2)	90 days, extendable to 180 days, Section 43D(2)	180 days
Bail test	Twin conditions, Section 20(8), no time relief	Not-guilty satisfaction, Section 49(7), statutory relief after one year	Prima facie true bar, Section 43D(5), no time relief	Prima facie bar, relieved only by constitutional courts for prolonged custody
Anticipatory bail	Barred, Section 20(7)	Barred, Section 49(5)	Barred, Section 43D(4)	Barred under all three designs
Presumption against the accused	Court shall presume, Section 21	Court shall draw adverse inference, Section 53	Court shall presume, Section 43E, terrorist act cases only	Presumption confined to Section 15 offences
Second definitional limb	Disruptive activity, Section 4	None	Unlawful activity, Section 2(1)(o)	Both limbs present in the UAPA scheme
Proscription of organisations	None	Schedule listing, Section 18	Tribunal-confirmed bans and Schedule listing, Sections 3 and 35	Two proscription tracks
Designation of individuals	No	No	Yes, from 2019, Section 35 and Fourth Schedule	Yes
Membership offence	None	Section 20, up to ten years	Sections 10 and 38, two and ten years	Yes, at two levels
Trial forum	Designated Courts, Section 9	Special Courts, Section 23	Sessions courts and NIA Special Courts	Ordinary and agency courts
Appeal	Supreme Court only, Section 19	High Court, Section 34	High Court	High Court
Witness identity protection	Section 16	Section 30	Section 44	Present in all three
Interception regime inside the Act	None	Chapter V, Sections 36 to 48	None	Interception sits in other statutes
Review of prosecutions	None	Review Committees, Section 60, binding from 2003	None	No prosecution review
Sunset or fixed duration	Two-year term, repeatedly renewed, lapsed 1995	Fixed three-year term, Section 1(6), repealed 2004	None	Permanent
Prior sanction for prosecution	Yes, Section 20A	Yes, Section 50	Yes, Section 45	Required under all three

Read vertically, the matrix shows why summary judgments fail. The TADA column is the most severe on custody and on the appeal route, and tied most severe on confessions. The POTA column is the most elaborately safeguarded, with a review mechanism and a bail valve that neither of the others has, while matching TADA on confessions. The UAPA column is the mildest on evidence and the broadest on reach, holding both definitional limbs, both proscription tracks and the only power to designate a named individual. No column is mildest throughout, and no column is harshest throughout, which is precisely the finding the political argument refuses to carry.

Three cautions govern the honest use of the matrix, and a reader who skips them will misuse it. The first is that the columns are not contemporaneous objects. The TADA column merges the Acts of 1985 and 1987, whose operative design was identical, and freezes them at lapse in 1995. The POTA column is the Act as it stood at repeal, including the 2003 amendment that armed its Review Committees, because that is the statute a revival would actually restore. The UAPA column is a composite built across fifty years, a 1967 base carrying chapters and sections inserted in 2004, 2008, 2013 and 2019, and no single Parliament ever enacted the column as a whole; anyone who finds that troubling has understood the article’s historical argument, which is that the composite was assembled precisely by never having to be defended as a whole. The second caution is that severity is measured against a moving baseline. Each statute modified the Code of Criminal Procedure and the Indian Evidence Act of its own day, and the ordinary law has itself been replaced since, so the comparison fixes its baseline deliberately: each anti-terror provision is measured against the ordinary criminal process it displaced at the time, which is the only comparison that captures what the legislature of the day chose to do. The third caution is that the matrix compares powers, not outcomes, and the section on application below states what the outcome record can and cannot add. A power unused is still a power: it shapes investigation and plea decisions whether or not it is ever read out in a judgment, and a comparison that counted only litigated provisions would understate every column.

How the powers moved: the repeal and transplant of 2004

The comparison has a hinge, and it is the legislative package of September 2004. POTA was not simply repealed and replaced by nothing. The Prevention of Terrorism (Repeal) Act and the Unlawful Activities (Prevention) Amendment Act were brought into force together on 21 September 2004, and the second of them rebuilt the UAPA using POTA as its principal quarry. The transplant can be traced provision by provision. The definition of a terrorist act in Section 15 of the UAPA follows the structure of POTA’s Section 3. The offences of membership, support and fundraising in Sections 38, 39 and 40 carry POTA’s Sections 20, 21 and 22 across, with the same maximum sentences of ten, ten and fourteen years. The listing of terrorist organisations in the First Schedule under Section 35 reproduces POTA’s Schedule mechanism, and the forfeiture chapter follows POTA’s scheme for the proceeds of terrorism. Even POTA’s Section 4 offence, unauthorised possession of arms and explosives in a notified area, survives in altered form, since unauthorised arms recovery is one of the triggers of the UAPA’s Section 43E presumption rather than an offence in itself. A reader who places the two texts side by side is not looking at two statutes that resemble each other by accident; the second was drafted out of the first.

The selectivity of the transplant is where the politics sits, and both sides of the 2004 argument should be given in their own terms. The United Progressive Alliance government, whose Common Minimum Programme had committed it to repealing POTA, presented the package as a considered separation: the offences and organisational machinery of the repealed Act were worth keeping, while its defining procedural departures, the police confession, the absence of any check on prosecutions, and the emergency form itself, were not. The Bharatiya Janata Party, which had enacted POTA and was by then in opposition, answered that the government had repealed the name and kept the substance while discarding the one provision, the confession rule, that made the substance enforceable in court, and that the result was the worst of both designs. Civil liberties organisations answered both from a third position: the People’s Union for Civil Liberties and allied groups argued that the transplant proved repeal to be cosmetic, since a citizen arrested after September 2004 faced custody, bail and presumption rules indistinguishable from POTA’s. The comparison discipline cuts through all three claims by forcing each to identify its row. The government’s claim is true of the confession and review rows. The opposition’s claim is true of nothing except emphasis, since the powers it valued most, listing, custody and bail, all survived. The civil liberties claim is true of the custody, bail and presumption rows and false of the evidence row. The transplant is the event that made the modern comparison possible, because it converted a succession of statutes into a single statute with a genealogy, and every later argument about whether the law in force is POTA under another name is an argument about this one legislative afternoon.

One feature of the transplant deserves separate notice, because it is easy to miss in the provision-by-provision account. POTA’s powers had been drafted as a self-contained code with its safeguards inside the same covers: the confession power sat a few sections away from its magistrate-production safeguard, and the Review Committee stood in the same Act as the offences it reviewed. The transplant moved the powers into a statute of 1967 whose architecture was built for a different subject, the control of unlawful associations, and left the safeguards behind in the repealed text. The result is a statute whose severe provisions and whose checking provisions do not face each other on the page, and in part do not exist at all. When later Parliaments amended the UAPA in 2008 and 2019, they added powers, custody rules and the designation of individuals; no amendment has ever revisited the missing half of the 2004 bargain. That asymmetry, addition without restoration, is the single most reliable predictor of how the comparison will look after the next amendment, whichever government moves it.

The second transplant: the reconstruction of 2008

The comparison has a second hinge, quieter than 2004 but nearly as important. The Mumbai attacks of November 2008 produced the predictable demand for POTA’s restoration, pressed by the opposition and by state governments, and the government’s answer defined the modern statute. Rather than revive the repealed Act or enact a new one, the United Progressive Alliance government, with P. Chidambaram as Home Minister, amended the UAPA again in December 2008 and created the National Investigation Agency alongside it. The 2008 amendment inserted the procedural core the 2004 transplant had omitted or left incomplete: Section 43D, carrying the ninety to one hundred and eighty day custody clock and the prima facie bail bar in sub-section (5); Section 43E, restoring the TADA-form presumption that POTA had softened into an adverse inference; and companion provisions on arrest, information and forfeiture procedure in Sections 43A to 43F. What 2008 did not insert is as significant for the comparison as what it did. The confession power was considered in the aftermath of an attack that killed over one hundred and sixty people, in a Parliament in no mood for tenderness, and it was still not restored. The prosecution Review Committee was not restored either, and no duration clause was added. The 2008 settlement therefore ratified the 2004 selection under maximum pressure: custody, bail and presumption were hardened to and beyond the POTA level, while the two features that had defined the earlier statutes’ reputations, the police confession and, on the safeguard side, the committee, stayed out. Any account of the UAPA as POTA revived has to survive this evidence, and it cannot, because the one provision that made POTA POTA in the public mind is the provision two successive Parliaments, in opposite political weather, declined to re-enact.

The 2008 moment also fixed the forum row in its modern form. By creating a federal investigative agency with its own Special Courts, the settlement gave the Union an institutional presence in terrorism investigation that neither TADA nor POTA had possessed, both of which had relied entirely on state police forces and state-constituted courts. The comparison of statutes understates this change if it reads texts alone, since the NIA Act is a fourth text standing behind the UAPA column of the matrix; but the institutional fact belongs in the comparison, because it altered what the UAPA’s provisions mean in practice for the cases the Centre chooses to take over, a category that has grown by notification ever since.

The confession row: where the verdict on wrongful conviction is decided

The single most important difference among the three statutes is the one the standard comparison usually gets wrong or omits. Under Section 25 of the Indian Evidence Act, a confession made to a police officer cannot be proved against the accused at all; the rule, explained in the guide to the Indian Evidence Act, rests on the judgment that a person in police custody speaks under a pressure the courtroom cannot later measure, so the law refuses the product altogether rather than attempting to filter it. TADA broke that rule. Section 15 of TADA made a confession recorded by a police officer not below the rank of Superintendent of Police admissible in the trial of the person who made it, and of a co-accused tried with him. POTA copied the device almost exactly: Section 32 of POTA made admissible a confession made before an officer not lower in rank than a Superintendent of Police, recorded in writing or on a mechanical or electronic device. The UAPA contains no equivalent provision at all. Nothing in the 2004 transplant, the 2008 reconstruction or any later amendment restored the device, so a confession to a police officer in a UAPA investigation is governed by the ordinary bar and is worthless to the prosecution as evidence of guilt.

Both statutes that carried the device wrapped it in safeguards, and the safeguards deserve recording because they define what revival would actually restore. The recording officer had to explain to the person in writing that he was not bound to confess and that a confession could be used against him, and could not compel or induce a confession from a person who preferred silence. The confession had to be recorded in an atmosphere free from threat or inducement and in the language in which it was made. The person then had to be produced before a Chief Judicial Magistrate or Chief Metropolitan Magistrate within forty eight hours with the original record, so that any complaint of torture could be made promptly and answered with a medical examination, after which the person went to judicial custody rather than back to the police. The Constitution Bench in Kartar Singh v State of Punjab upheld the TADA provision in 1994 subject to strict compliance with safeguards of this kind, treating the seniority of the recording officer and the speed of magisterial production as the conditions on which admissibility depended. The Supreme Court in People’s Union for Civil Liberties v Union of India, decided in December 2003, upheld the POTA provision on the same footing.

Set against those safeguards is the structural objection that persuaded Parliament in 2004 to leave the device behind, and it has never been answered on its own terms. A safeguard polices the recording of a confession; it cannot police the hours of custody that precede the recording, which is when the pressure the Evidence Act worries about is applied. Critics of the device, among them the People’s Union for Civil Liberties in its challenge to POTA and Human Rights Watch in its reporting on TADA, made the further point that admissibility changes police incentives: where a signed confession can carry a prosecution, obtaining one becomes a substitute for investigating, and the cases that result are strong on paper and weak everywhere else. The defence of the device, stated at full strength by the governments that enacted it and accepted in substance by the Kartar Singh majority, is that terrorism cases are distinctive because witnesses are killed or terrified into silence, so the ordinary evidentiary rule, written for crimes committed in public, leaves the prosecution with nothing where the offence was planned in private; a confession recorded by a senior officer under magisterial supervision was said to be the proportionate answer. A comparison article does not have to resolve that argument to score the row. On any view, the power to prove a police confession is a power that operates directly on the risk of wrongful conviction, since a false confession, once admitted, is the hardest evidence in a trial to displace. TADA had it, POTA had it, the UAPA does not, and that single absence is why describing the statute in force as simply the harshest of the three is false.

The device also outlived its statutes in the courtroom, which is part of what a comparison must capture. Because the savings in TADA and the Repeal Act preserved pending prosecutions, trials founded on Section 15 and Section 32 confessions continued for years after 1995 and 2004, and the largest terrorism trials of the period, including the trials arising out of the Bombay blasts of 1993, were conducted under TADA and rested substantially on confessions recorded by senior police officers, tested against the Kartar Singh safeguards. A confession-led prosecution has a distinctive shape that the comparison should name: the investigation converges early on the recorded statement, the trial becomes an argument about the circumstances of the recording rather than about the underlying events, and the safeguards, the warning, the forty eight hour production, the medical examination on a complaint of torture, become the real battleground. Where the safeguards were strictly proved, convictions followed; where they were not, the confession fell and often the case with it. The UAPA prosecution has no equivalent centre of gravity. Its trials are built from recoveries, electronic records, financial trails and witness testimony assembled in the ordinary way, which is one reason they are slower and one reason their failures look different: a UAPA case collapses for want of proof, where a TADA case collapsed for abuse of the shortcut. Neither failure mode is attractive, but they are not the same failure, and the difference is the whole of the confession row stated in operational terms.

The custody clock: one year, six months, six months

The second row runs the other way, and it is longer than most summaries admit for the oldest statute. All three laws rewrote Section 167 of the Code of Criminal Procedure, the provision that entitles an accused to release when the investigation is not completed within a fixed period. TADA, by Section 20(4), stretched that period furthest: the investigation was to be completed within one hundred and eighty days, and the Designated Court could extend the period up to one year on the report of the Public Prosecutor showing the progress of the investigation and the specific reasons for detaining the accused beyond one hundred and eighty days. A person held for a TADA offence could therefore spend a full year in custody, lawfully, before the state had to commit itself to a charge sheet or release him. POTA, by Section 49(2), halved that outer limit: ninety days, extendable by the Special Court to one hundred and eighty days on the same kind of Public Prosecutor’s report. The UAPA, by Section 43D(2), adopted the POTA figures exactly, ninety days extendable to one hundred and eighty, and has kept them unchanged through every later amendment.

The comparison worth making is not only among the three statutes but between them and the ordinary law they displaced. Under the unmodified Code, the ceiling was ninety days for the gravest offences and sixty for the rest, so even the mildest of the three anti-terror regimes doubled the ordinary maximum, and the TADA regime quadrupled it at the outer limit. The doubling matters in a specific way that aggregate talk of harshness misses. Pre-charge custody is the period in which the accused is most completely in the state’s hands and least able to prepare a defence, and its length is set entirely by the statute, without any judicial finding about guilt. On this row TADA stands alone as the most severe, and the two later statutes are indistinguishable. A proposal to revive TADA would therefore add six months of lawful pre-charge detention to the statute book, a consequence that revival debates almost never mention, since the debate fastens on confessions and forgets the clock.

At the end of each clock sits the same legal event with three different dates: default bail. When the period expires without a charge sheet, the accused acquires what the case law calls an indefeasible right to be released on bail on furnishing the bond the court directs, provided the right is availed of before the charge sheet is filed. The doctrine was worked out under TADA itself, in Hitendra Vishnu Thakur, where the Supreme Court insisted the accused must be told of the prosecutor’s report seeking extension, and in Sanjay Dutt v State (1994), where a Constitution Bench construed the right arising under Section 20(4) and held that it must be exercised when it accrues, since a charge sheet filed thereafter extinguishes it. The practical meaning of each statute’s clock is therefore set by two numbers, not one: the first date, at which the prosecution must either file or formally seek more time in a hearing the accused can contest, and the final date, at which silence becomes release. Under the ordinary law those dates are sixty or ninety days, TADA stretched them to one hundred and eighty and three hundred and sixty-five, and POTA and the UAPA settled at ninety and one hundred and eighty. The default-bail event also explains the rhythm of real cases under all three statutes, the flurry of charge sheets filed on the eve of expiry, the extension applications argued like miniature trials, and the occasional release of a serious accused on a technicality that the prosecution’s own delay created. Custody clocks are usually discussed as grants of power to the state; their other face, visible in this row, is that every extension of the clock is also a postponement of the day the state’s failure to finish becomes the accused’s right to leave.

Bail: three formulations of the same refusal, with one valve

Bail is the row on which the comparison becomes genuinely fine-grained, because all three statutes refuse bail in a formula, and the formulas differ in ways that decide real applications. TADA, by Section 20(8), imposed the twin conditions: where the Public Prosecutor opposed the application, no person in custody was to be released unless the court was satisfied that there were reasonable grounds for believing that he was not guilty of the offence and that he was not likely to commit any offence while on bail. The double requirement is what made the provision so difficult to satisfy, since an accused who could raise a real doubt about guilt could still be held on a prediction about his future conduct, and the provision contained no relief with the passage of time. TADA also excluded anticipatory bail altogether, by Section 20(7), as did POTA by Section 49(5) and the UAPA by Section 43D(4); on that point the three statutes never differed.

POTA’s bail provision, Section 49, looked similar and was structurally kinder. Sub-section (6) required the Public Prosecutor to be heard. Sub-section (7) provided that where the prosecutor opposed release, the accused was not to be released until the court was satisfied that there were grounds for believing that he was not guilty of the offence: a single condition, without TADA’s second limb about future offences. Then came the feature no other Indian anti-terror statute has carried. A proviso to sub-section (7) provided that after the expiry of one year from the date of detention, only sub-section (6) would apply, so the not-guilty satisfaction fell away and bail returned to ordinary judicial consideration with the prosecutor heard. POTA, in other words, accepted that a statutory presumption against bail could be tolerated for a year and not indefinitely, and wrote that limit into the Act itself. The UAPA’s provision, Section 43D(5), is differently built and in operation the strictest of the three. It bars release, for offences under Chapters IV and VI of the Act, where the court, on a perusal of the case diary or the police report, is of the opinion that there are reasonable grounds for believing that the accusation is prima facie true. The question the provision asks is not whether the accused can show grounds for believing himself not guilty, but whether the prosecution’s own papers, taken at face value, disclose a case. In National Investigation Agency v Zahoor Ahmad Shah Watali, the Supreme Court held that at the bail stage the court must proceed on the prosecution material as it stands and may not weigh it as a trial court would, which converts the prima facie test into something close to a presumption of detention wherever a charge sheet has been competently drafted; the judicial interpretation of UAPA bail works through that case law in full, and this article does not re-answer it.

The UAPA contains no equivalent of POTA’s one-year valve. The relief that exists has come from constitutional adjudication rather than from the statute: in Union of India v K. A. Najeeb, the Supreme Court held that a constitutional court may grant bail despite Section 43D(5) where prolonged incarceration without prospect of a speedy trial offends the right to life and personal liberty under Article 21. The comparative point is sharp. Under TADA, bail was hardest to obtain at the start and stayed hardest. Under POTA, the statute itself relaxed after a year. Under the UAPA, the bar is set by the prosecution’s papers and relaxes only when a constitutional court is persuaded that the length of the detention has become a constitutional wrong, which is a remedy measured in years and available in few cases. Which formulation is strictest depends on the month of custody in which the question is asked: in the first year, TADA’s twin conditions and the UAPA’s prima facie bar are both close to conclusive, with the UAPA’s test the more mechanical; beyond the first year, POTA’s accused had a statutory route out that neither of the others’ accused has ever had.

One further bail feature separates TADA from both successors and is easily overlooked because it concerns a small class of accused. POTA, by Section 49(9), and the UAPA, by Section 43D(7), each provide that no bail shall be granted to a person accused under the Act who is not an Indian citizen and has entered the country unauthorisedly or illegally, except in very exceptional circumstances and for reasons recorded in writing. TADA contained no citizenship condition of any kind: its bail bar turned entirely on the twin conditions, whatever the accused’s nationality. The later statutes thus added a status-based exclusion that the earliest and otherwise harshest of the three never enacted, a reminder that the direction of travel on individual rows is not uniform even inside a single subject. For the small number of foreign accused in infiltration cases, the practical consequence is that the bail inquiry under the two later laws begins from a statutory near-prohibition that no TADA court was directed to apply.

How a case begins: arrest, investigation and the first days

The comparison usually starts at the charge sheet, but the statutes diverge from the first hour. POTA is the only one of the three that wrote arrest discipline into its own text. Section 52 required the arresting officer to prepare a custody memo of the person arrested, to inform him on arrival at the police station of his right to consult a legal practitioner, to communicate the arrest immediately to a family member or relative by telegram, telephone or other means and to record that communication under the arrested person’s signature, and to permit the arrested person to meet his legal practitioner during interrogation, though not to have the practitioner present throughout it. Section 51 fixed a seniority floor for the investigation itself: no officer below the rank of Deputy Superintendent of Police, or Assistant Commissioner in the metropolitan areas, could investigate a POTA offence. Neither TADA nor the UAPA contains equivalents. TADA’s front end was governed by the Code, supplemented by the one distinctive feature already noticed, the requirement of prior approval of the Director General of Police before information of an offence was recorded at all, a control on entry into the statute rather than on the treatment of the person arrested. The UAPA likewise leaves arrest and investigation to the ordinary law, to the Code’s requirements and to the constitutional safeguards the Supreme Court built in cases such as D. K. Basu v State of West Bengal, which bind every arrest in India whatever statute is invoked.

The row matters because the first days of an anti-terror case are when its character is set: the decision to invoke the special statute, taken at the police station, determines the custody clock, the bail test and the forum for everything that follows, and it is taken before any court sees the file. POTA’s drafters, whatever the Act’s reputation, are the only ones who legislated for that moment in the statute itself, custody memo, family intimation, lawyer access and a seniority floor for the investigator. The UAPA investigator works under no statutory rank floor and no statutory arrest code, and the TADA investigator worked under a prior-approval regime that controlled the file’s opening but not the suspect’s treatment. When the comparison turns from the courtroom to the police station, the ordering of the three statutes is not the ordering the headlines suggest, and reform proposals drawn from the comparison should notice that the most detailed front-end code in the field belongs to the statute that was repealed for its harshness.

Membership’s two levels inside the statutes

The proscription section recorded that POTA and the UAPA punish membership and TADA did not, but each of the two also grades the offence internally, and the grading reveals different theories of what membership means. POTA carried two membership offences of very different weight. Section 3(5) punished membership of a terrorist gang or a terrorist organisation involved in terrorist acts with imprisonment up to life, treating the member of an active violent body as standing close to the acts themselves. Section 20 punished belonging or professing to belong to an organisation listed in the Schedule with up to ten years, treating membership of the named body as the offence regardless of any particular act. The UAPA’s ladder is built on a different distinction, the legal character of the body rather than its involvement in acts: two years under Section 10 for membership of an association declared unlawful through the Tribunal-confirmed process, and ten years under Section 38 for membership of an organisation listed as a terrorist organisation by executive notification. The practical difference is considerable. Under POTA’s upper limb, the prosecution had to connect the organisation to terrorist acts to reach the life maximum. Under the UAPA’s Section 38, the listing notification does that work in advance, and the trial is about belonging. It was this feature that made the membership offence the most litigated of the organisation provisions, read down in Arup Bhuyan on its Section 10 sibling and restored in 2023, and it is the point at which the modern statute’s design choice, liability by designation, is most exposed.

The same facts through three statutes

Formulations are abstract until they are run against a fact pattern, and the comparison sharpens considerably when the same three recurring situations are passed through each statute in turn. Take first the speech case: a public figure addresses a meeting and praises a banned organisation, and nothing else is alleged. Under TADA, if the speech questioned the sovereignty or territorial integrity of India, it could be charged as a disruptive activity under Section 4, an offence with a five-year minimum, triable only by a Designated Court, with bail governed by the twin conditions and any appeal lying to the Supreme Court; the Vaiko prosecution under POTA is the real instance of the pattern in the second statute, where a speech in support of the Liberation Tigers of Tamil Eelam was prosecuted as support for a terrorist organisation under Section 21, and where the Review Committee machinery eventually became the route by which the case was scrutinised outside the prosecution’s own hierarchy. Under the UAPA the same speech is analysed in two tracks: as support for a terrorist organisation under Section 39 if the body is listed in the First Schedule, and as an unlawful activity under Sections 2(1)(o) and 13 if it questions sovereignty or supports secession, with the association itself liable to be declared unlawful. The pattern shows what the matrix cannot: the speech case is chargeable under all three statutes, but the forum, the bail test and the existence of any external review differ at every step, so two accused persons a decade apart, on identical facts, faced materially different legal jeopardy.

Take second the slow investigation: a conspiracy whose evidence sits in foreign bank records and intercepted communications abroad, so that the investigating agency cannot complete its work quickly. The custody clock decides the case before the trial begins. TADA gave the agency one hundred and eighty days and, on a prosecutor’s report, up to a year, and the Supreme Court in Hitendra Vishnu Thakur v State of Maharashtra policed the extension power by insisting that the accused be told of the report and that the right to default bail, once accrued and availed of, could not be defeated by a charge sheet filed afterwards. POTA and the UAPA give ninety days and a single extension to one hundred and eighty on the same kind of report, and in both the extension hearing is the first real contest of the case. The difference of six months between the TADA ceiling and the shared later ceiling is the difference between a prosecution that can wait out a foreign evidence request inside one remand and a prosecution that must either file an incomplete charge sheet or watch the accused walk out on default bail; investigators who worked under both regimes are consistent in identifying the year-long clock, not the confession rule, as the TADA power they felt most in daily work.

Take third the confession-led case: the agency’s strongest material is what the accused said in custody. Under TADA and POTA the statement, if recorded by an officer of the required rank with the safeguards proved, could be placed before the court and could found a conviction, including against a co-accused tried jointly, which made the recording the centre of the investigation. Under the UAPA the same statement is inadmissible, and its only lawful uses are the narrow ones the ordinary law allows, principally the discovery of facts in consequence of information received, under the proviso tradition of the evidence law; the investigation must therefore be built to stand without it. The pattern isolates the largest single operational difference among the three statutes: two of them permitted an investigation to terminate in a signature, and one requires it to terminate in a record.

Presumptions: the quiet row, and a difference in drafting that matters

Each statute relieved the prosecution of part of its burden, and the row repays close reading because the three texts do not say the same thing. Section 21 of TADA directed the Designated Court to presume, unless the contrary was proved, that the accused had committed the offence, where it was proved that arms or explosives specified in the Act were recovered from his possession with reason to believe that similar material had been used in the offence, or that his fingerprints were found at the site or on an object used in it; a parallel limb covered the accused who had rendered financial assistance to a person accused or reasonably suspected of a connected offence. Section 53 of POTA covered the same two triggers, recovery of arms or explosives and fingerprints at the scene, and added a financial-assistance limb of its own, but chose a different operative phrase: the Special Court was to draw an adverse inference against the accused. Section 43E of the UAPA, inserted by the 2008 amendment, returned to the TADA formulation, directing the court to presume, unless the contrary is shown, that the accused committed the offence under Section 15, on the same recovery and fingerprint triggers and none other.

The difference between presuming an offence and drawing an adverse inference is not verbal decoration. A presumption in the TADA and UAPA form places the burden of displacing the presumed fact on the accused once the trigger facts are proved, and the presumed fact is the offence itself. An adverse inference invites the court to reason against the accused from the proved facts but leaves the legal burden where the ordinary law puts it. Commentators on POTA at the time of its enactment treated the softer phrase as a deliberate dilution, and the dilution survived the PUCL challenge without judicial conversion into the stronger form. The reach of the UAPA presumption is meanwhile narrower than TADA’s in one respect worth noting: Section 43E is confined to prosecutions for the terrorist act offence under Section 15, whereas the TADA presumption sat across the terrorist act offences of its Section 3 and the financial-assistance presumption reached conduct charged under sub-section (3) of that section. On this row the order of severity runs TADA first, the UAPA second, POTA third, an ordering that cuts against the assumption that each successor simply intensified its predecessor.

Definitions: what counts as terrorism, and the limb each statute added

The definitions decide who enters the regime at all, and here the comparison has a twist that the matrix records but the prose must explain. The core definition of a terrorist act is broadly stable across the three. TADA’s Section 3 addressed acts done with intent to overawe the government or to strike terror in the people, by means of bombs, dynamite, firearms and similar instruments, causing or likely to cause death, injury, property damage or disruption of essential supplies and services. POTA’s Section 3(1) widened the means to acts done by any other means whatsoever, and listed the intents as threatening the unity, integrity, security or sovereignty of India or striking terror. The UAPA’s Section 15, inserted in 2004 and expanded since, follows the POTA structure and adds two extensions of its own: the threat to economic security was written into the definition, and the 2012 amendment, enacted as Act 3 of 2013, brought acts threatening the monetary stability of India, including by high-quality counterfeit currency, within it. On the core definition, then, the sequence is one of steady widening, and the statute in force has the widest text of the three.

The difference lies in the second limb each statute carried beside the terrorist act. TADA’s second limb was the disruptive activity of Section 4, which criminalised action, including speech, that questioned or supported a claim questioning the sovereignty and territorial integrity of India, punishable with a minimum of five years’ imprisonment; it was this limb, detachable from any act of violence, that produced many of the most criticised TADA prosecutions, since it converted advocacy about the map of India into an offence in the same statute as bombing. POTA dropped the second limb entirely: it had one target, the terrorist act and its associated offences, and nothing corresponding to disruptive activity. The UAPA, however, was never only an anti-terror statute, and its original subject supplies a second limb of a different construction. Section 2(1)(o) defines unlawful activity to include action, including spoken or written words, that supports a claim to cession or secession or that disclaims, questions or disrupts the sovereignty and territorial integrity of India, and that definition feeds the Act’s association offences and its machinery against unlawful associations. The practical consequence is that the speech-adjacent offence TADA was condemned for, and POTA abandoned, survives in the statute in force in a parallel track, punishable under Section 13 with imprisonment of up to seven years, even though it sits outside the terrorist act chapter and outside the Section 43D(5) bail bar. A comparison that only reads the terrorism chapters of the three Acts will miss this entirely.

Definitional width has a second consequence that operates before any trial: it hands the choice of regime to the prosecution. Where the same conduct, an explosion, a speech, a transfer of money, can be charged under the ordinary Penal Code or under the special statute, the decision to invoke the special statute is the decision that fixes the custody clock, the bail test, the presumption and the forum, and that decision is taken by the investigating agency subject to sanction, not by a court. TADA’s history is the demonstration. The disruptive activity limb was drawn widely enough to cover conduct far from any insurgency, and the statute’s use spread accordingly, to states and to categories of crime its drafters had not aimed at, because width at the definition stage converts directly into discretion at the charging stage. POTA’s narrower single-limb design was in part a reaction to that experience, though its terrorist act definition was itself wide in the means it covered. The UAPA combines the widest terrorist act text of the three with a second limb of TADA-like breadth, so the invocation discretion in the law in force is at least as great as it ever was under TADA, and it is exercised under a bail regime less forgiving than TADA’s. Readers who rank the statutes by their definitions alone will therefore reach a different order from readers who rank by evidence, and both are measuring something real: the definition decides how many people the rest of the machine can reach.

Proscription and designation: the row where the survivor outgrew both predecessors

On the machinery for striking at organisations, the order of the three statutes reverses the usual story of decline. TADA had no proscription machinery of any kind. It created no list of banned organisations, no offence of membership as such, and no tribunal or committee to review a listing, because its design was aimed at acts and at the individuals who committed them; a person could be prosecuted under TADA for what he did, for harbouring, for conspiracy, but not for the bare fact of belonging to a named organisation, since the statute named none. POTA introduced the list. Its Chapter III allowed the Central Government to add an organisation to the Schedule as a terrorist organisation, created by Section 20 an offence of belonging or professing to belong to a listed organisation, punishable with up to ten years’ imprisonment, and added offences of support and of fundraising in Sections 21 and 22, punishable with up to ten and fourteen years. Section 19 gave the organisation or an affected person an application for removal from the Schedule, with review by the Review Committee if the government refused. The Supreme Court in the PUCL case read a requirement of intention or knowledge into Sections 20 to 22, so that the offences punished a knowing association rather than an accidental one.

The UAPA took the POTA machinery, kept it, and added a second track beside it, and then in 2019 did something neither predecessor attempted. The Act’s older machinery, in Chapter III, allows the Central Government to declare an association unlawful, a declaration that has no effect until confirmed by a Tribunal presided over by a sitting High Court judge, which must decide within six months of the notification; membership of an unlawful association is an offence under Section 10 punishable with up to two years, a provision the Supreme Court read down in Arup Bhuyan in 2011 to require incitement to violence and then, on review in 2023, upheld in its plain terms. Beside that track sits the POTA inheritance: Section 35 empowers the Central Government to list an organisation as a terrorist organisation in the First Schedule, with removal sought by application under Section 36 and review by a Review Committee under Section 37, and Sections 38, 39 and 40 punish membership, support and fundraising at up to ten, ten and fourteen years respectively, matching the POTA figures. The 2019 amendment then extended Section 35 to individuals, creating a Fourth Schedule of named persons designated as terrorists, a power no earlier Indian anti-terror statute had contained: TADA designated nobody, POTA listed organisations only, and the UAPA alone can attach the legal label of terrorist to a single named person by executive notification, with the consequences for reputation, property scrutiny and prosecution that follow, and with no prior hearing and no confirming tribunal in the design. Any honest statement of which statute is the most far-reaching has to record that the newest power in the field belongs exclusively to the oldest statute.

Money: proceeds and forfeiture across the three statutes

The financial perimeter of the three statutes shows the same pattern of late expansion. TADA’s property provisions were the thinnest of the three: the statute concentrated on the person of the accused, and such forfeiture as it provided operated at the end of the process, against the property of a person convicted, rather than as an investigative instrument available from the first day. POTA rebuilt the subject entirely. It made the holding of proceeds of terrorism illegal in itself by Section 6, and by Section 7 armed an investigating officer not below the rank of Superintendent of Police, with the prior written approval of the Director General of Police, to seize or attach property believed to represent such proceeds, subject to confirmation or revocation by the Designated Authority after hearing the person affected. The design is recognisably modern: freeze first, adjudicate the freezing quickly, and let the criminal trial proceed separately. The UAPA took the scheme over in the 2004 transplant and keeps a forfeiture chapter on the same model, administered through a Designated Authority, running in parallel with the prosecution. The comparative point is double-edged in the usual way of this field. On reach, the financial machinery of the law in force is far beyond anything TADA possessed, and it bites on bank accounts, property and intermediaries who may never face a charge themselves. On discipline, the machinery carries its own judicial and quasi-judicial checkpoints of a kind the prosecution side of the Act lacks, so that the money row is simultaneously the broadest and one of the better supervised parts of the modern statute. Arguments that treat the UAPA as uniformly less checked than its predecessors have to reckon with the fact that its property regime, inherited from POTA, contains a confirmation hearing that its bail regime has never had.

Forum and appeal: the row TADA lost outright

All three statutes distrusted the ordinary trial process, but they distrusted it to different depths. TADA created Designated Courts under Section 9, presided over by judges appointed with the concurrence of the Chief Justice of the High Court, with power to sit in camera and to try connected offences, and then, by Section 19, removed the High Court from the appellate structure altogether: an appeal lay as a matter of right from a Designated Court to the Supreme Court alone, on facts and on law, and no other appeal or revision lay anywhere. The design bought speed at the price of a tier, and it placed the whole appellate burden of a national statute on a single court; the Supreme Court in Kartar Singh noted the strain and directed that vacancies in Designated Courts be filled so that undertrials did not languish. POTA corrected the design. Its Special Courts under Section 23 were staffed the same way, but Section 34 sent appeals to a bench of two judges of the High Court, both on facts and on law, restoring the intermediate tier that TADA had abolished. The UAPA created no special court of its own at all: its offences are tried by the ordinary Court of Session, and where the National Investigation Agency investigates, by the Special Courts constituted under the National Investigation Agency Act 2008, with appeals lying to the High Court in the ordinary way. On forum and appeal, therefore, the sequence is not an intensification but a normalisation: the exceptional court with the truncated appeal existed only under TADA, was softened under POTA, and survives under the UAPA only in the agency’s cases.

The witness provisions tell the same story of convergence rather than escalation. Each statute empowered its courts to keep the identity and address of a witness secret and to hold proceedings in camera, TADA by Section 16, POTA by Section 30, and the UAPA by Section 44, and each made breach of a protective direction an offence. The provisions answer a genuine and continuing problem, the exposure of prosecution witnesses to retaliation in terrorism trials, and on this row there is nothing to choose among the three texts; the honest comparison records a power that all three legislatures, across four decades and three political dispensations, independently decided a terrorism trial cannot do without.

One trial power of POTA’s deserves separate record because it has no counterpart in the statute in force and is rarely mentioned in the comparison. Section 29(5) of POTA permitted a Special Court, for reasons recorded, to proceed with the trial in the absence of the accused or his pleader and to record the evidence of a witness, subject to the accused’s right to recall the witness for cross-examination. Trial in absence sits at the edge of what a fair-trial guarantee can bear, and its presence in POTA complicates the statute’s reputation as the safeguarded member of the family: the same Act that wrote the custody memo and the Review Committee also authorised evidence to be taken behind the accused’s back in a defined circumstance. The UAPA contains no such power, and trials under it follow the Code’s insistence on the accused’s presence, subject to the ordinary exceptions. The row is a useful corrective in both directions, against romanticising POTA’s safeguards and against treating the modern statute as the maximum of every severity the field has known.

The state dimension: the confession power’s second life

A comparison confined to central statutes would leave one large fact outside the frame: the device that disappeared from central law in 2004 did not disappear from Indian law. The Maharashtra Control of Organised Crime Act 1999, a state statute aimed at organised crime syndicates rather than at terrorism as such, carried the template’s signature features into state law, including, in Section 18, the admissibility of confessions recorded by an officer not below the rank of Deputy Commissioner of Police, and a bail provision modelled on the twin-condition pattern. Karnataka and other states enacted cognate organised crime statutes on similar lines. The Supreme Court upheld the Maharashtra Act’s design, and confession-led prosecutions under it have been a continuing feature of organised crime trials long after the last POTA confession was recorded. The comparison therefore has a jurisdictional footnote that qualifies its headline verdict. At the level of the Union’s anti-terror code, the confession power is gone and the evidence row is settled. At the level of Indian criminal law as a whole, the power survives in the adjacent field of organised crime, available against overlapping conduct, since the same network can present as a terrorist organisation to one agency and as a crime syndicate to another. Legislators who congratulated themselves in 2004 on retiring the device retired it from one statute book among many, and a complete account of the three central laws should record where it went.

The state dimension cuts the other way on custody and bail as well. State organised crime statutes adopted the extended remand and restrictive bail features of the central template, often in terms as severe as anything in the UAPA, so the liberty rows of the comparison describe not a central peculiarity but a national drafting habit that reproduced itself at both levels of the federation. What is distinctive about the central comparison remains what it was: only the central statutes carry the proscription and designation machinery, only they define the terrorist act for the whole country, and only the UAPA among them is permanent. But the reader who generalises from the matrix to Indian anti-crime law entire should know that its mildest finding, the death of the police confession, is a finding about three named statutes and not about the legal system that contains them.

Interception and review: the machinery around the case

Two rows of the matrix concern not the trial but the apparatus around it, and on both the POTA column is the anomaly. POTA was the only one of the three statutes to contain its own interception code. Chapter V, running from Section 36 to Section 48, allowed an investigating officer not below the rank of Superintendent of Police to apply to a Competent Authority, a senior secretary-level officer, for authorisation to intercept wire, electronic or oral communication, on stated grounds of probable cause, for not more than sixty days at a time; the authorisation then had to be placed before the Review Committee within seven days for approval. TADA contained nothing of the kind, and the UAPA contains nothing of the kind either: interception in a UAPA investigation takes place under the Indian Telegraph Act 1885 and the Information Technology Act 2000, statutes of general application with their own authorisation machinery, and the anti-terror statute neither expands nor specifically regulates it. Commentators who opposed the 2004 transplant made the point from the other side as well, noting that powers taken from POTA into the UAPA arrived without the POTA safeguards that had surrounded them; the interception example shows the pattern precisely, since there the power was exercised under general law both before and after, and only during POTA’s short life was it codified inside the anti-terror statute with a dedicated reviewing body.

The review row is the comparison’s clearest safeguard gap. POTA, by Section 60, required the Central and State Governments to constitute Review Committees chaired by a person who is or has been a High Court judge, and an amendment deemed in force from 27 October 2003 inserted sub-sections (4) to (7) into that section, empowering a committee, on the application of an aggrieved person, to review whether a prima facie case existed for proceeding against an accused, and making its directions binding on the government and the investigating officer, with proceedings deemed withdrawn where the committee found no prima facie case. When POTA was repealed, the Repeal Act of 2004 extended the same review to every pending POTA case, whether or not anyone applied, and required it to be completed within a year. Neither TADA nor the UAPA has ever contained a mechanism by which a body outside the prosecution can order a prosecution reviewed and, if it finds nothing in it, end it. The Constitution Bench in Kartar Singh, confronting the same structural worry under TADA, recommended screening committees as an administrative safeguard, but recommendation is not enactment, and the TADA text carried no such body. The UAPA’s Tribunal and its Section 37 Review Committee both concern listings, of associations and organisations, not prosecutions; there is no UAPA body to which an accused person can say that the case against him should never have been brought. A legislature reviving one feature of POTA would find more safeguard per section in this committee than in any other part of the repealed Act.

Accountability for misuse: sanction, compensation and the absent remedy

A comparison of powers is incomplete without a comparison of the brakes fitted to them, and here the three statutes share more weakness than any of them admits. All three required the previous sanction of government before a court could take cognizance of an offence, Section 50 of POTA and Section 45 of the UAPA in terms, while TADA went further in Section 20A, requiring the prior approval of the Director General of Police before information about an offence was even recorded, and government sanction before cognizance. In design, sanction is the point at which a political superior takes responsibility for a prosecution; in operation, across all three regimes, it functioned as a filter that caught almost nothing, since the sanctioning authority and the prosecuting authority answer to the same government. The comparison records the row because the failure is uniform: no one of the three statutes can claim a working sanction filter as a point of superiority.

Beyond sanction, the statutes diverge in one small and telling place. POTA alone punished its own misuse. Section 58 made a police officer who exercised powers corruptly or maliciously, knowing there were no reasonable grounds for proceeding, punishable with imprisonment of up to two years, and empowered the Special Court to award compensation to a person corruptly or maliciously proceeded against, payable by the officer or authority responsible. Neither TADA nor the UAPA contains any equivalent: an officer who invokes the 1967 Act without grounds faces departmental consequences and the ordinary law, but the statute itself provides the victim no compensation route and the officer no specific penalty. The Constitution Bench in Kartar Singh saw the gap in TADA and tried to fill it administratively, directing that screening or review committees be constituted to examine TADA cases before and during prosecution, a judicial recommendation that governments implemented unevenly and that bound no one. The National Human Rights Commission’s criticism of TADA’s operation, recorded in the renewal debates, pressed the same point from outside the courts. The result across the three statutes is a pattern the matrix understates: the powers were legislated in detail and the remedies for their abuse were left to implication, recommendation and the general law, in 1987, in 2002 and in every UAPA amendment since. A reader comparing safeguard design should therefore score POTA’s Section 58 as the field’s only enacted experiment in statutory self-discipline, small as it was, and its disappearance in 2004, unmourned by either side of the repeal debate, as evidence of how low the accountability row ranks in the politics of all three laws.

Time limits: two designs of sunset, and none

Because the brief for this comparison is often summarised with the claim that only TADA had a sunset clause, the record should be stated exactly. TADA’s time limit worked by renewal: enacted for two years, extended by amendment three times, and extinguished by non-renewal in 1995, so that the statute’s death required no repealing Act and left pending prosecutions to be carried on under its savings. POTA’s time limit worked by fixed term: Section 1(6) provided that the Act, deemed in force from 24 October 2001, was to remain in force for three years, with savings for proceedings begun under it; the repeal of September 2004 in fact reached the statute about a month before its own term expired, so POTA holds the unusual distinction of having been both term-limited and repealed. The difference between the designs is not cosmetic. A renewal-based sunset forces the government of the day to seek fresh authority and to defend the record, which is what happened, with rising difficulty, in 1989, 1991 and 1993. A fixed-term sunset requires nothing of anyone until the term runs out, and can be overtaken by events, as POTA’s was. The UAPA has neither: passed as permanent legislation in 1967, it has no duration clause, no renewal point and no scheduled moment at which Parliament must ask whether its anti-terror powers should continue. Arguments for restoring a sunset to the law in force can borrow from both precedents, but they are arguments for different institutions: the TADA design disciplines by repetition, the POTA design by a single deadline, and the experience of the two suggests the discipline lies in the repetition rather than in the deadline. The point cuts both ways, and the other side should be stated fairly. A renewal requirement hands a future Parliament a recurring opportunity to let a needed law die in a bad session, and TADA’s own history shows renewals being granted on thin scrutiny in 1989 and 1991 before the discipline bit in 1995; a sunset is a ratchet that can slip. The defenders of permanence answer further that the UAPA’s objectionable features can be amended at any time, so the absence of a sunset is a failure of political will rather than of design. The comparison cannot settle that institutional argument, but it can report the scoreboard: in the absence of any expiry mechanism, the statute has been amended four times, on each occasion to add power, and never to subtract it, while both expiry designs, whatever their theoretical weaknesses, did in fact bring their statutes to an end.

The courts’ three encounters, and the one that never happened

Each of the temporary statutes was carried to the Supreme Court and tested as a whole, and the pattern of those tests is itself a finding of the comparison. TADA met its test in Kartar Singh v State of Punjab, where a five-judge Constitution Bench, in a judgment delivered on 11 March 1994 and reported at (1994) 3 SCC 569, upheld the Act across nearly its whole extent: the definitions, the confession provision subject to safeguards, the bail conditions, the exclusion of anticipatory bail and the Designated Court scheme all survived. The one provision struck down was Section 22, which had made a person identifiable from a photograph in circumstances the Court found unintelligible and opposed to the fair procedure guaranteed by Article 21. The template the case set, uphold the statute, discipline its operation with safeguards, remove the single indefensible section, was followed nine years later when POTA met its test in People’s Union for Civil Liberties v Union of India, decided on 16 December 2003 and reported at (2004) 9 SCC 580, which upheld POTA, read intention and knowledge into its organisation offences, and insisted on the safeguards built around the confession provision. In neither case did comprehensive challenge produce comprehensive invalidation; what it produced was a judicially annotated statute and a public record of how the power was to be exercised.

Why has the UAPA never faced a whole-Act challenge of the Kartar Singh kind?

Because challenges to it have been brought provision by provision rather than against the Act as a whole: the membership offence was tested in Arup Bhuyan, the bail bar in Watali and Najeeb, and the 2019 designation power is under challenge in petitions directed at that amendment, so no bench has ever been asked to assess the statute entire.

The asymmetry deserves emphasis, because it inverts the assumption that survival implies judicial approval. The two statutes that received a comprehensive constitutional audit are gone, and the audit is part of the public record that made their renewal indefensible. The statute that survives has been examined closely at several points, its bail provision authoritatively glossed in Watali, its membership provision read down and then restored on review in Arup Bhuyan, and its designation amendment awaiting adjudication, but it has never stood in the dock as a whole, and no court has ever pronounced on the cumulative weight of its custody, bail, presumption, proscription and designation powers in a single case. That is not a defect in the litigation; it is a consequence of the statute’s biography. A temporary law concentrates opposition into a single challenge mounted while renewal is being debated. A permanent law fought in pieces produces doctrine in pieces, and the pieces never assemble into the earlier verdicts’ whole-statute scrutiny.

The pattern across the three encounters also taught the field a strategic lesson that shapes how the statute in force is resisted. Two comprehensive challenges produced two comprehensive legitimations: Kartar Singh left TADA standing with one section struck down, and the PUCL judgment left POTA standing with meanings supplied. Challenges framed against single provisions, by contrast, have repeatedly extracted concessions without conferring legitimacy on the whole: the reading down of the membership offence in 2011 stood for twelve years, the Najeeb route opened a constitutional bypass around the bail bar without striking the bar down, and the Tribunals under the proscription chapters have been held to real standards of disclosure and reason-giving in individual cases. Advocates who absorbed the lesson resist the UAPA the way it was built, provision by provision and amendment by amendment, which means the piecemeal scrutiny the statute receives is not an accident of litigation but the rational response to the precedent the two whole-Act challenges created. The judicial row of the comparison should therefore be read with its feedback loop attached: the earlier statutes were audited whole because they were enacted whole, and their audits immunised the technique of enactment itself, while the statute assembled in pieces escapes the whole-Act audit and is resisted in pieces for the same reason.

Text against application: the objection that the comparison is the wrong exercise

The most serious objection to everything above is that it compares texts while the harm lives in application. A statute’s severity in the hands of a police force that books first and investigates later, the objection runs, is set less by its sections than by its invocation: the same custody clock is a different instrument in a district where it is used in a handful of cases and in one where it is used in hundreds, and a comparison of formulations mistakes the map for the territory. The objection has real force, and it is the reason this series gives the enforcement record its own article rather than folding it into a table; the enforcement record of the anti-terror laws examines arrest patterns, acquittals and the human residue of all three regimes, and this article does not duplicate it. What the comparison of texts can still do, and what application talk cannot do without it, is fix the ceiling of what application may lawfully become. Application varies by state, by decade and by government; the text is the part that a legislature chose and the only part a repeal or an amendment can change.

The figures that are available, stated with their sources because conviction numbers in this field are contested and are routinely quoted without any, support the objection’s premise that outcomes under all three regimes were dominated by acquittal and abandonment rather than conviction. Human Rights Watch recorded that TADA produced over 76,000 arrests with a conviction rate of under one per cent. For POTA, the count assembled by the journalist Jyoti Sethi in the study published as Kafkaland recorded 1,031 arrests, of which only 18 cases had completed trial and 13 ended in conviction within the period studied; and when POTA’s pending cases were subjected to the Repeal Act’s general review after 2004, a large share was found to disclose no prima facie case at all and was withdrawn. For the UAPA, the Ministry of Home Affairs told the Rajya Sabha, on the basis of the National Crime Records Bureau’s 2019 Crime in India report, that 2.2 per cent of the cases registered under the Act between 2016 and 2019 ended in conviction. Three regimes, three different texts, and one recurring shape of outcome, which is why the application objection cannot be waved away. The sociologist of criminal process Malcolm Feeley gave the underlying mechanism its enduring name in his 1979 study The Process Is the Punishment: where the pre-trial process itself imposes years of custody, lost livelihood and stigma, the formal outcome at the end of it, conviction or acquittal, ceases to be the main event, and the sanction has already been served by the person the state failed to convict. Feeley’s subject was a lower court in Connecticut, but his analysis is the standard account of why the custody and bail rows of the matrix matter more to the lived severity of the three statutes than their penalty rows, on which all three agree. It is also why the application objection, taken seriously, does not dissolve the textual comparison but reorders it: measured by what the process does to the unconvicted, the rows that decide severity are the clock and the bail bar, and on those rows the statute in force is the equal or the superior in harshness of both its predecessors. It is also why the textual comparison remains worth doing: if outcomes converge while texts diverge, then the marginal contribution of any single provision can be isolated, and the provisions that widen the gap between arrest and conviction, the custody clock and the bail bar, can be identified as the ones doing the work.

Journalists who cover these trials work inside that gap between arrest and outcome for years at a stretch, holding the names of sources who spoke about an investigation on condition they never surface in it, and the craft of that work is its own subject, well described in how journalists use VaultBook to protect sources while reporting trials that run for years.

Does the verdict change if enforcement, rather than the statutory text, is compared?

Partly. On outcomes, the three regimes converge towards mass arrest and rare conviction, which softens every textual distinction except two: the confession power, which changes what a conviction rests on where one is obtained, and the bail and custody provisions, which decide how the years before acquittal are spent.

The ordinary law standing behind all three

No anti-terror prosecution is ever only an anti-terror prosecution. Each of the three statutes layered its regime on top of the Indian Penal Code, the Arms Act, the Explosive Substances Act and the evidence and procedure codes, and a single investigation characteristically produces charges under the special Act and under the ordinary law together. The layering has a comparative consequence that single-statute accounts miss: when the special charge fails, the ordinary charges frequently remain, so the practical severity of a regime is partly determined by how much of the case survives the special Act’s collapse. The best-known illustration is also the most litigated TADA prosecution of all. Sanjay Dutt, tried for his part in the network behind the Bombay blasts of 1993, was acquitted by the Designated Court of the TADA charges, including conspiracy, and convicted instead under the Arms Act for unauthorised possession of prohibited weapons, a conviction the Supreme Court upheld in 2013. The special statute had governed his custody, his bail and his forum for over a decade of proceedings, and the conviction that finally stood was an ordinary-law conviction. The pattern generalises across the three regimes: the special Act supplies the investigation’s frame, its clock and its bail discipline, while the substantive conviction, where one is obtained at all, very often rests on offences any police force could have charged without it. A comparison of the three statutes that ignored the ordinary law behind them would overstate what each added; the addition was real but specific, concentrated in procedure, evidence and organisational liability, and it is exactly those concentrated additions that the matrix isolates.

What a revival bill would have to do

The demand to bring back TADA or POTA is a recurring feature of the aftermath of major attacks, and the comparison gives it a precise content it usually lacks. A revival bill would not be filling a vacuum. It would be adding, to a statute book that already contains the offences, the listing machinery, the custody clock and the bail bar, the small residue of powers that 2004 and the lapse of 1995 actually removed: the admissibility of a confession recorded by a senior police officer, TADA’s one-year pre-charge ceiling, TADA’s appeal to the Supreme Court alone, and, if POTA is the model, an interception code and a prosecution Review Committee. A drafter would also have to decide the question both temporary statutes answered and the UAPA never has: the duration of the new law, and with it whether the discipline of renewal or of a fixed term is to be restored. And the drafter would know, from the record this article has set out, exactly what reception awaits the bill in the Supreme Court: the Kartar Singh and PUCL template, under which a comprehensive challenge produces a comprehensive audit, the statute upheld in substance, its safeguards converted into conditions of validity, and any indefensible section struck down. Revival, in other words, would purchase one evidentiary power at the price of the whole-statute scrutiny the law in force has never undergone. That exchange is the real content of the revival demand, and it is rarely how the demand is presented.

There is also an asymmetry in what legislation can reach. A revival Act operates prospectively; it cannot disturb the acquittals, the withdrawn prosecutions or the concluded trials of the earlier regimes, and it cannot shorten the detention of anyone held under the UAPA in the meantime. The cases that revival rhetoric usually invokes, the acquittals that followed long trials under the repealed laws, belong to the savings provisions and to history. What a legislature can change is only the future balance among the rows of the matrix, which is why the honest form of the revival argument is not a name but a list: restore the confession power, or restore the custody year, or restore the review committee, each defended on its own row. Framed that way, some revival proposals become harder to sustain and others, the committee above all, become harder to refuse, which is the service a power-by-power comparison exists to perform.

The verdict, and who should do what with it

The verdict can be stated with the precision the matrix earns. On evidence, the statute in force is milder than both predecessors, by the largest single margin in the comparison: it cannot prove a confession made to a police officer, and TADA and POTA both could. On liberty before trial, it is more severe than POTA, whose custody ceiling it matches while lacking POTA’s one-year bail valve and prosecution Review Committee, and it is more severe than TADA on the quality of the bail test while less severe on the custody clock, since TADA alone allowed a year before charge. On reach, it is broader than either: it alone carries both a disruptive-activity analogue and a terrorist act definition, both proscription tracks, and the power to designate individuals. On accountability, it is the least supervised of the three, having neither a sunset of either design nor a review body for prosecutions, and it is the only one never audited as a whole by a constitutional bench. “Milder on evidence, harsher on liberty” is the short form of that verdict, and its deciding factor is the one named at the outset: the reader who counts a wrongful conviction as the worst thing a legal system can do will rank the UAPA best of the three, and the reader who counts years of detention ending in acquittal as the worst will rank it worst, and both rankings are honest readings of the same matrix.

What is the deciding factor when one anti-terror law is called harsher than another?

The deciding factor is which harm the comparison is built to prevent. Ranked by risk of wrongful conviction, the confession row dominates and the UAPA is the mildest of the three; ranked by risk of prolonged detention without conviction, the custody and bail rows dominate and the UAPA is the most severe in force.

For the legislator or the staffer drafting in this field, the matrix functions as a menu with prices attached. Restoring a repealed statute would add remarkably little that is not already law: of the powers in the TADA and POTA columns, only the police confession, TADA’s year-long custody ceiling and TADA’s direct Supreme Court appeal are genuinely absent from the statute in force, and each of those is a power whose restoration would reverse the direction the law has travelled since 2004. The reforms that would change the comparison most cost nothing in investigative reach: a statutory bail valve on the POTA model, a prosecution review committee on the Section 60 model, and a renewal-based duration clause would each move the UAPA column without touching a single offence definition. For the advocate, the comparison is a working tool rather than an argument. An old TADA prosecution still in the system is governed by TADA’s text, its twin bail conditions and its direct appeal, not by anything the UAPA says; a POTA matter saved at repeal carries the Review Committee history with it; and a UAPA case turns on the prima facie discipline of Watali and the constitutional opening left by Najeeb. It should be recorded once, in the place where practical readers will look for it, that this comparison is general information about three statutes and not advice on any individual case, and that in this field limitation periods and forum rules are unforgiving of delay. For the aspirant, the comparison is a corrective to the textbook progression: the examined answer that scores is the one that separates the rows, gives the confession row to the predecessors, the liberty rows against the statute in force, and refuses a single ranking until the criterion is named.

Which statute works as the template and which as the warning for a future drafting committee?

POTA works as the template for safeguards, since it alone combined its powers with a fixed term, a one-year bail valve and a binding prosecution review; TADA works as the warning, since its renewal decade shows how a temporary statute with the widest custody clock and a speech limb becomes indefensible on its own record.

Three classes of case stand outside the general verdict, and honesty requires naming them. The first is the legacy prosecution: a trial saved from the TADA or POTA era is governed by the repealed text in full, so for the accused in such a case, some of whom have faced proceedings spanning decades, the comparison is not historical at all and the harsher column is the live one. The second is the organisational case, where the modern statute’s advantages to the prosecution are greatest: an accused charged through the listing machinery, with membership and support offences, a designation notification in evidence and the Section 43D(5) bar applied to the organisation chapters, faces a cumulative structure neither predecessor assembled, and for that accused the UAPA is the most severe of the three on almost every row that touches him. The third is the investigation built on intercepted communications, where POTA, paradoxically, offered the subject of surveillance a codified authorisation process and a reviewing committee inside the same statute, while the law in force routes interception through general statutes whose safeguards are real but were not designed for terrorism investigations and are not argued in the trial. In each class the verdict’s split character survives, but its centre of gravity moves, and a reader who knows which class a live question belongs to will use the matrix better than a reader who memorises its headline.

Frequently Asked Questions

Q: What is the difference between TADA, POTA and the UAPA?

The three are India’s successive central anti-terror statutes, and they differ power by power rather than in overall design. TADA, in force from 1985 and renewed until its lapse in 1995, admitted confessions recorded by senior police officers, allowed up to a year of custody before a charge sheet, and sent appeals directly to the Supreme Court. POTA, in force from 2001 to 2004, kept the confession power but halved the custody ceiling, restored appeals to the High Court, and added a prosecution Review Committee and a three-year fixed term. The UAPA, permanent legislation of 1967 expanded in 2004, 2008, 2013 and 2019, dropped the confession power entirely, kept the shorter custody ceiling and the strictest bail test of the three, and added proscription of organisations and designation of individuals that neither predecessor combined.

Q: Which of the three anti-terror laws was the harshest?

There is no single answer, because the ranking reverses by row. TADA was harshest on custody, allowing one hundred and eighty days before charge extendable to a full year, and on appeal, which lay only to the Supreme Court. TADA and POTA tie as harshest on evidence, since both made a confession to a police officer admissible, the power most closely connected to wrongful conviction. The UAPA is harshest on bail, where the prosecution’s papers need only appear prima facie true, and on reach, since it alone designates individuals. A reader who fears wrongful conviction most will name TADA or POTA; a reader who fears endless pre-trial detention most will name the UAPA.

Q: Why did TADA and POTA lapse while the UAPA survived?

The difference is structural rather than a judgment on merits. TADA was temporary legislation that continued only if Parliament renewed it, and in 1995 the government declined to seek renewal after a decade of criticism of its record. POTA carried a fixed three-year term and was in any event repealed in September 2004 by a newly elected governing coalition. The UAPA was enacted in 1967 as permanent legislation with no duration clause, so its continuation never required a vote, a defence of its record, or a renewal debate. It survived because ending it would have required positive legislative action, while keeping it required none, and because each time a temporary statute fell, the powers considered indispensable were amended into it.

Q: Is a confession to a police officer admissible under TADA, POTA or the UAPA?

Under TADA and POTA, yes; under the UAPA, no. Section 15 of TADA and Section 32 of POTA each made a confession recorded by a police officer not below the rank of Superintendent of Police admissible at trial, subject to safeguards including a written warning that the person was not bound to confess, recording free of threat or inducement, and production before a magistrate within forty eight hours. The UAPA contains no such provision, so the ordinary rule of evidence law applies and a confession made to a police officer cannot be proved against the accused. This single difference is the largest point of contrast in the whole comparison, because it decides what a conviction may rest on.

Q: Which anti-terror law had the strictest bail provision?

Measured by the time over which the bar operates, the UAPA’s provision is the strictest. TADA’s Section 20(8) imposed twin conditions, reasonable grounds for believing the accused not guilty and not likely to offend on bail, which were extremely difficult to satisfy but were at least addressed to the accused’s case. POTA’s Section 49 required the court to be satisfied of grounds for believing the accused not guilty, but a proviso released that requirement after one year of detention. The UAPA’s Section 43D(5) asks only whether the prosecution’s own case diary or report discloses reasonable grounds for believing the accusation prima facie true, contains no time-based relief, and has been read in the Watali judgment to forbid weighing the material as a trial court would.

Q: Did the UAPA absorb the provisions of the repealed laws?

Substantially, yes, though selectively. When POTA was repealed in 2004, the same legislative exercise amended the UAPA to insert a chapter of terrorist act offences modelled on POTA, including the definition in Section 15, offences of membership, support and fundraising, and the listing of terrorist organisations in a schedule. The 2008 amendment added the extended custody period, the prima facie bail bar and the presumption in Section 43E. What was not absorbed is as significant: the police confession power, the prosecution Review Committee and every form of time limit were left behind. The UAPA is therefore best understood as the accumulated powers of the earlier statutes, minus their two most contested evidence and safeguard features, resting on a permanent base enacted in 1967.

Q: Which anti-terror law had a sunset clause and why does that matter?

Both temporary statutes had one, in different designs, and the UAPA has none. TADA was enacted for two years and continued only through successive renewing amendments in 1989, 1991 and 1993, so its survival repeatedly required Parliament to defend its record until renewal was refused in 1995. POTA carried a fixed term instead: Section 1(6) limited it to three years from its commencement in October 2001, and it was repealed about a month before that term expired. The distinction matters because a renewal-based sunset forces periodic justification, while a fixed term merely sets a deadline. The UAPA, permanent since 1967, has never faced either discipline, which is central to why it outlived both.

Q: Would reviving POTA add anything to current powers?

Very little, and the addition can be named exactly. Almost every operative POTA power already exists in the UAPA in the same or stronger form: the terrorist act offences, organisation listing, membership and fundraising offences, extended custody and restrictive bail all survive in the amended 1967 Act. A revived POTA would genuinely add only three things: the admissibility of confessions recorded by senior police officers under Section 32, which the UAPA deliberately omits; the interception code of its Chapter V, which operates under general statutes instead; and, on the safeguard side, the prosecution Review Committee of Section 60 and the one-year relaxation of its bail bar. Revival debates that ignore this inventory mistake a famous name for a missing power.

Q: How do the definitions of a terrorist act differ across TADA, POTA and the UAPA?

The core is stable and the edges move. TADA’s Section 3 required intent to overawe the government or strike terror, through bombs, dynamite, firearms or similar means, causing death, injury, damage or disruption of essential services. POTA’s Section 3(1) listed the intents as threatening the unity, integrity, security or sovereignty of India or striking terror, and widened the means to any other means whatsoever. The UAPA’s Section 15 follows POTA’s structure and extends it further, adding threats to economic security and, by the 2012 amendment, acts threatening India’s monetary stability such as high-quality counterfeit currency. Beside the core, TADA separately punished disruptive activity and the UAPA separately defines unlawful activity, while POTA had no second limb.

Q: Which of the three laws allowed the longest detention before a charge sheet had to be filed?

TADA, by a wide margin. Section 20(4) of TADA gave the investigation one hundred and eighty days, and empowered the Designated Court, on the Public Prosecutor’s report showing progress and specific reasons, to extend the period up to one year before the accused became entitled to release for default. POTA’s Section 49(2) allowed ninety days, extendable by the Special Court to one hundred and eighty. The UAPA’s Section 43D(2) adopted the POTA figures exactly, ninety days extendable to one hundred and eighty, and no later amendment has altered them. The ordinary Code of Criminal Procedure ceiling of ninety days for the gravest offences puts all three in perspective: even the shortest anti-terror clock doubled it, and TADA’s quadrupled it.

Q: How did the presumptions against an accused differ under TADA, POTA and the UAPA?

The trigger facts are nearly identical and the legal consequence is not. Under Section 21 of TADA, where arms or explosives were recovered from the accused with reason to believe similar material was used in the offence, or fingerprints were found at the site, the Designated Court was directed to presume, unless the contrary was proved, that the accused had committed the offence; a matching limb covered financial assistance. Section 53 of POTA used the same triggers but directed the court only to draw an adverse inference, which leaves the burden of proof formally with the prosecution. Section 43E of the UAPA restored the stronger TADA form, the court shall presume unless the contrary is shown, but confined it to prosecutions for the terrorist act offence under Section 15.

Q: Which of the three laws empowered the government to proscribe an organisation, and by what machinery?

POTA and the UAPA, but not TADA. TADA contained no list of organisations and no banning machinery at all; it attached to acts, not to bodies. POTA’s Chapter III allowed the Central Government to enter an organisation in the Schedule as a terrorist organisation, with an application for removal under Section 19 and review by the Review Committee. The UAPA runs two tracks: under Chapter III an association may be declared unlawful, a declaration requiring confirmation by a Tribunal presided over by a sitting High Court judge, which must decide within six months of the notification, and under Section 35 a terrorist organisation may be listed in the First Schedule, with removal by application under Section 36 and review under Section 37. Since 2019 the second track also reaches named individuals.

Q: How did the three laws treat the raising of funds for terrorism?

TADA had no free-standing fundraising offence; money appeared in it mainly through the presumption in Section 21, which treated proved financial assistance to a person accused or suspected of a connected offence as ground for presuming the offence itself. POTA created the offence directly: Section 22 punished inviting, receiving or providing money or property intended or suspected for the purposes of terrorism with imprisonment of up to fourteen years. The UAPA split the subject in two. Section 17 punishes raising funds for a terrorist act itself with imprisonment from five years to life, while Section 40 punishes raising funds for a terrorist organisation with up to fourteen years, tracking the POTA figure for the organisational offence.

Q: How did the trial forum and appeal route differ under the three laws?

TADA was the most exceptional. Its Designated Courts, constituted under Section 9, tried its offences exclusively, and Section 19 made an appeal lie as a matter of right to the Supreme Court alone, on facts and on law, excluding the High Courts from the structure entirely. POTA constituted Special Courts under Section 23 but restored the intermediate tier, Section 34 sending appeals to a bench of two judges of the High Court. The UAPA created no court of its own: its offences are tried by the ordinary Court of Session, or by Special Courts under the National Investigation Agency Act where that agency investigates, and appeals follow the ordinary route to the High Court. The sequence runs from exceptional to ordinary, against the trend of most other rows.

Q: Which of the three laws contained a wiretapping regime, and what became of it?

Only POTA. Its Chapter V, Sections 36 to 48, built a complete interception code inside the anti-terror statute: an officer not below the rank of Superintendent of Police applied to a Competent Authority of secretary rank, which could authorise interception of wire, electronic or oral communication for up to sixty days at a time on stated probable cause, and every authorisation had to be placed before the Review Committee within seven days for approval. TADA contained no interception provisions, and neither does the UAPA. When POTA was repealed, the code disappeared with it, and interception in terrorism investigations returned to the general machinery of the Indian Telegraph Act 1885 and the Information Technology Act 2000, where it has remained.

Q: How did the constitutional challenges to the three laws compare in outcome?

The two comprehensive challenges produced the same template. In Kartar Singh v State of Punjab, a five-judge Constitution Bench in 1994 upheld TADA almost entire, including the confession provision subject to safeguards, striking down only Section 22 on identification by photograph. In People’s Union for Civil Liberties v Union of India, decided in December 2003, the Supreme Court upheld POTA, reading intention and knowledge into its organisation offences. The UAPA has never been tested as a whole. Its provisions have been challenged and construed one at a time, membership in Arup Bhuyan, bail in Watali and Najeeb, and the 2019 designation power in petitions directed at that amendment, so no bench has assessed the accumulated statute in a single case.

Q: Which of the three laws made membership of a proscribed organisation an offence by itself?

POTA and the UAPA, and under TADA the question cannot arise because it proscribed nothing. Section 20 of POTA punished belonging or professing to belong to a listed terrorist organisation with imprisonment of up to ten years, and the PUCL judgment read in a requirement of intention or knowledge. The UAPA carries the offence at two levels: membership of an association declared unlawful is punishable under Section 10 with up to two years, and membership of a listed terrorist organisation under Section 38 with up to ten years, the figure POTA had used. The Supreme Court read Section 10 down in Arup Bhuyan in 2011 to require more than passive membership, and a review bench in 2023 upheld the provision in its plain terms.

Q: What penalties did the three laws prescribe for committing a terrorist act?

On this row the three statutes converge almost completely. TADA’s Section 3(2), POTA’s Section 3(2) and the UAPA’s Section 16 each prescribe death or imprisonment for life where the terrorist act results in death. Where it does not, each prescribes imprisonment for a term that shall not be less than five years and may extend to imprisonment for life, together with liability to fine. Conspiracy, attempt, abetment and preparation sit at the same five-year floor in each scheme. The convergence is instructive for the comparison as a whole: the legislatures that disagreed about confessions, custody, bail and duration never disagreed about the punishment for the completed act, which is why severity rankings built on penalties alone find nothing to separate the three laws.

Q: Which law built in a review committee for prosecutions, and what could that committee do?

Only POTA. Section 60 required the Central and State Governments to constitute Review Committees chaired by a serving or retired High Court judge. As enacted, the committees advised on listing and interception matters, but an amendment deemed in force from 27 October 2003 inserted sub-sections (4) to (7), under which a committee could, on the application of an aggrieved person, review whether a prima facie case existed for proceeding against an accused, issue directions binding on the government and the investigating officer, and, where it found no prima facie case, cause the proceedings to be deemed withdrawn. The 2004 Repeal Act then applied the same review to every pending POTA case. Neither TADA nor the UAPA has ever given an outside body power to end a prosecution.

Q: How did the three laws protect the identity of witnesses?

In materially the same words, because each legislature copied the solution rather than the problem. TADA by Section 16, POTA by Section 30 and the UAPA by Section 44 each empower the trial court to hold proceedings in camera and to take measures for keeping the identity and address of a witness secret, including avoiding mention of names in orders and judgments and directing that identity not be disclosed, with breach of a direction punishable as an offence. The protection answers the feature of terrorism trials that all three statutes confronted, the exposure of witnesses to intimidation and retaliation, and it is the one procedural power on which the comparison finds complete agreement across forty years of legislating.