Introduction – Section 45 At The Crossroads
For over two decades, Section 45 of the Prevention of Money Laundering Act, 2002 (“PMLA”) has occupied an uneasy position in Indian criminal jurisprudence. It imposes stringent “twin conditions” for bail in money laundering cases, while Article 21 of the Constitution of India guarantees that no person shall be deprived of personal liberty except according to a just, fair and reasonable procedure. When pre-trial incarceration under Section 45 is prolonged and trials move at a glacial pace, this tension becomes acute. The Hon’ble Supreme Court has, at different points, struck down Section 45 in part, upheld its amended form, and then crafted Article 21 based carve outs for cases of excessive incarceration. Against this backdrop, a recent decision of the Punjab and Haryana High Court in Varun Puri v. Directorate of Enforcement[1] advances a striking conceptual move wherein it treats Article 21 as the “Grundnorm” of the criminal process, before which Section 45 must yield when the State fails its correlative obligation to ensure a fair and expeditious trial.
The Twin Conditions And Their Original Design
Section 45 of PMLA as originally enacted, imposed two preconditions before any court could grant bail in respect of a PMLA offence:
(i) The Public Prosecutor had to be given an opportunity to oppose the bail application; and
(ii) Where such opposition was made, the court had to be satisfied that there were reasonable grounds for believing that the accused was not guilty of the offence of money laundering and was not likely to commit any offence while on bail.
This formulation was drawn almost verbatim from provisions such as Section 37 of the Narcotic Drugs and Psychotropic Substances Act, 1985, and Section 20(8) of the Terrorist and Disruptive Activities (Prevention) Act, 1987, both of which were enacted for offences involving grave threats to public order and national security. However, the original Section 45 contained a structural anomaly wherein its application was tethered not to the money laundering offence itself, but to whether the predicate (scheduled) offence attracted a sentence of more than three years. As a result, similarly placed accused persons could face radically different bail regimes depending solely on the nature of the scheduled offence. An accused could show, at the bail stage, that there were reasonable grounds for believing that he was not guilty of money laundering, yet remain subject to the rigours of Section 45 merely because of the scheduled offence’s sentencing range. Liberty under the PMLA thus became contingent on the attributes of an offence over which the PMLA court had no primary jurisdiction.[2]
The Article 21 Problem: Presumption Of Innocence And Prolonged Custody
At a deeper level, Section 45’s requirement that the court be satisfied of “non-guilt” before trial placed significant pressure on the presumption of innocence, a principle that the Supreme Court has characterised as lying at the heart of Article 21. When combined with systemic trial delays, the twin conditions risk converting the presumption of innocence into a presumption of continuing incarceration, an accused may remain in custody for an extended period, required to effectively demonstrate innocence at the bail stage, even while the state fails to progress the trial with reasonable dispatch. This is the constitutional dilemma that runs through later jurisprudence: can a reverse-burden bail clause coexist with prolonged pre-trial incarceration without violating Article 21’s guarantee of fair and reasonable procedure?
Doctrinal Evolution: From Structural Anomaly To Substantive Liberty Concerns
In Nikesh Tarachand Shah v. Union of India[3], a two-judge bench of the Supreme Court considered challenges to the original Section 45. The Court struck down Section 45(1) as unconstitutional, identifying multiple defects/grounds:
(i) Anomalous consequences flowing from the schedule-based trigger;
(ii) Lack of rational nexus between the bail restriction and the object of the PMLA;
(iii) Inconsistency between the treatment of anticipatory and regular bail; and
(iv) Reversal of the presumption of innocence inherent in Article 21.
While the first two grounds were anchored in Article 14, targeting the irrational classification created by the schedule-link, the third and fourth ground went to the core of Article 21. The Court noted that the same accused, for the same offence, could face different thresholds at different stages (anticipatory versus regular bail), without persuasive justification.
The 2018 Amendment: Textual Repair, Limited Doctrinal Retreat
Parliament responded through the Finance Act, 2018, substituting the schedule-link with the phrase “under this Act”, thereby tethering the twin conditions directly to the money laundering offence. The Finance Act (No. 2) of 2019 further introduced an Explanation providing that all PMLA offences are cognisable and non-bailable. This was presented as a legislative cure to the defects identified in Nikesh Tarachand Shah. In substance, however, the amendment addressed primarily the Article 14 related anomalies (irrational classifications and lack of rational nexus) by removing the dependence on the scheduled offence. It did not engage with the deeper Article 21 concerns flagged in Grounds C and D of Nikesh Tarachand Shah, namely: the inconsistent treatment of anticipatory versus regular bail and the inversion of the presumption of innocence. The core concern that Section 45 demanded a form of pre-trial self-exoneration remained embedded within the statutory text, now without the earlier schedule based absurdity.
A Formal Validity Yet Unresolved Tension
In Vijay Madanlal Choudhary v. Union of India[4], a three-judge bench upheld the constitutional validity of the amended Section 45. The Court stressed the perceived gravity of money laundering as a socio-economic offence undermining the nation’s financial system and relied on precedents upholding similar twin conditions in other special statutes, including the Constitution Bench decision in Kartar Singh v. State of Punjab, (1994) 3 SCC 569. It emphasised that Section 45 preserved a measure of judicial discretion and did not amount to an absolute bar on bail.
Yet, the structural critique persists. The comparison drawn in Kartar Singh between TADA and Section 437 of the Code of Criminal Procedure (now Section 480 id BNSS) rested on a standard that required the court only to see whether there “appear” reasonable grounds. Section 45 PMLA, like Section 20(8) TADA, requires the court to be “satisfied”, a more demanding epistemic threshold that operates in the inverse direction (non-guilt rather than guilt). In this sense, the Court in Vijay Madanlal restored formal validity to the amended provision but did not finally resolve the Article 21 concerns captured in Nikesh Tarachand Shah. Review petitions, now listed before a reconstituted bench, continue to raise the reversal of the presumption of innocence as a central issue. [5]
The Supreme Court’S Emerging Pattern
After Vijay Madanlal, a series of Supreme Court decisions signalled an Article 21 based counter current in the context of prolonged pre-trial incarceration under PMLA and analogous special statutes. In Manish Sisodia (II) v. Directorate of Enforcement[6], the Court granted bail after approximately 17 months of custody, observing that Section 45 would not stand in the way where trial delays were indefinite and incarceration had already been protracted. The Court candidly noted an institutional tendency of trial courts and High Courts to “play safe” in bail matters, sometimes in derogation of the principle that bail is the rule and jail the exception. [7]
Through a consistent line of precedents, the Supreme Court has established that while Section 45 of the Prevention of Money Laundering Act, 2002 remains formally valid, Article 21 operates as an overriding constitutional constraint against prolonged pre-trial incarceration and stalled trials. In V. Senthil Balaji v. Deputy Director, Directorate of Enforcement[8], the Court articulated the “correlative obligation principle” holding that stringent statutory thresholds for bail impose a reciprocal duty upon the State to ensure a speedy trial, as onerous statutory restrictions and indefinite trial delays cannot constitutionally coexist. This constitutional primacy was reinforced in Prem Prakash v. Union of India[9], where the Court affirmed that Section 45 does not overturn the foundational principle that “bail is the rule and jail is the exception” , and in Arvind Dham v. Directorate of Enforcement[10], which clarified that economic offences cannot be treated as a monolith to justify mechanical detention where trial progress is stalled.
Framing Of Article 21 As The Criminal Justice Grundnorm
It is within this evolving doctrinal landscape that the Punjab and Haryana High Court’s recent decision in Varun Puri v. Directorate of Enforcement[11], assumes significance. The judgment opens with Ulpian’s maxim that prison is for custody, not punishment, and draws upon Clause 40 of Magna Carta (“To no one will we sell, to no one will we refuse or delay, right or justice”) as well as Blackstone’s warning that confinement without swift trial is a dangerous engine of arbitrary government. These references are deployed to support a central proposition – the right against prolonged pre-trial incarceration is not a creature of statute but a pre-constitutional liberty that the Constitution recognises and codifies through Article 21. [12]
On this philosophical foundation, the Court advances its core doctrinal move in three steps. First, it treats Article 21 as the Grundnorm of the criminal process, an overarching constitutional norm to which all criminal statutes, including PMLA, remain subordinate in operation. Second, it locates the justification for Section 45’s twin conditions in the State’s corresponding duty to ensure that trials are conducted fairly and expeditiously. Third, it concludes that when the State fails in this duty, continued incarceration can no longer draw legitimacy solely from Section 45, instead, bail in such circumstances flows as a direct, non-negotiable right from Article 21.
Significantly, the Court also reads Section 479 of the Bharatiya Nagarik Suraksha Sanhita, 2023 (formerly Section 436A CRPC) as a constitutional floor rather than a ceiling. Where the progress of a trial makes a timely verdict effectively illusory, courts must not treat the statutory time thresholds as the outer limit of protection. Article 21, as Grundnorm, may require earlier intervention to secure liberty even before those statutory limits are technically met.
Reimagining Section 45’S Validity In Practice
The distinctive contribution of Varun Puri lies less in its outcome than in its restructuring of the doctrine governing Section 45. Earlier Supreme Court decisions had already recognised, in practice, that Article 21 could override Section 45 in cases of prolonged custody, but they did so largely on a case to case basis, without a fully articulated theoretical account. Varun Puri supplies that account. By casting Article 21 as the Grundnorm and making the continued legitimacy of Section 45 contingent upon the State’s performance of its speedy trial obligation, the judgment places statute and Constitution in a relationship of mutual conditionality. Section 45 remains formally valid, as held by the Supreme Court in Vijay Madanlal, but becomes inapplicable in concrete cases where the State has defaulted on its constitutional duty to prosecute with reasonable expedition. In that sense, the Article 21 override is not an exception grafted onto Section 45, it is a recognition that Section 45 never had autonomous authority to justify indefinite pre-trial incarceration in the first place.
The judgment also engages with empirical assessments of PMLA bail outcomes, which commentators have criticised as reflecting a “calibrated” rather than consistently constitutional approach. By articulating clear criteria for when and how Article 21 must prevail over Section 45, the judgement supplies defence counsel and courts with a more coherent doctrinal lens than case to case, ad hoc balancing between liberty and enforcement. As a single-judge High Court decision, it is not formally binding on other High Courts or on the Supreme Court, however, its persuasive force lies instead in the precision with which it advances a constitutional theory of bail under special statutes. The pending review petitions in Vijay Madanlal Choudhary, which are now listed before a bench headed by the Chief Justice of India, squarely raise the presumption of innocence concern first crystallised in Nikesh Tarachand Shah, namely that Section 45 inverts that presumption at the bail stage. The Grundnorm analysis in Varun Puri aligns directly with this objection and may offer the Supreme Court an attractive doctrinal template if it chooses to recalibrate the operation of Section 45 in cases of prolonged pre-trial incarceration.
Implications For Future Pmla Jurisprudence
The evolving jurisprudential trajectory from Nikesh Tarachand Shah through Vijay Madanlal, and subsequently Manish Sisodia (II), Senthil Balaji, Prem Prakash, Arvind Dham and Varun Puri, reveals an important trajectory. Section 45’s formal validity has been reaffirmed, but its capacity to justify prolonged pre-trial incarceration has been progressively narrowed by the reassertion of Article 21. The more clearly courts articulate the State’s correlative obligation to ensure a speedy, fair trial, the less room remains for treating the twin conditions as an impregnable barrier to bail in cases of sustained delay.
By explicitly naming Article 21 as the Grundnorm and treating statutory thresholds such as Section 479 BNSS as a minimum protective floor, the court crystallises this trend with unusual philosophical clarity. It confirms that the legitimacy of restrictive bail clauses in special statutes is not static but conditional on their operation within a constitutional framework that prioritises liberty over indefinite trial-less incarceration. Blackstone’s warning that indefinite confinement without trial is a subtle and dangerous engine of arbitrary government thus acquires renewed salience in Indian money laundering jurisprudence. Whether, and to what extent, the Supreme Court will adopt or modify this Grundnorm approach in the pending review of Vijay Madanlal will determine how far Section 45 can continue to coexist with the Constitution’s evolving understanding of personal liberty.
CONCLUSION
The evolution of Section 45 PMLA from Nikesh Tarachand Shah, the 2018 textual amendments and the Supreme Court’s reaffirmation of validity in Vijay Madanlal Choudhary, has produced a regime in which the provision is no longer assessed in isolation, but against the constitutional baseline of Article 21. Subsequent bail decisions of the Supreme Court in complex economic and special statute prosecutions have reinforced that even the most stringent twin-condition clauses cannot constitutionally sustain a prolonged, trial-less incarceration, and that the presumption of innocence, effective defence and the right to a speedy trial remain governing principles rather than dispensable aspirations. High Courts, too, have increasingly read Section 45 in the shadow of these guarantees, using delay, the nature of the evidence and the individualised risk of misuse of liberty as decisive factors in granting bail despite the statutory rigour. Against this backdrop, the pending review of Vijay Madanlal offers the Supreme Court an opportunity not so much to reopen Section 45’s bare text as to clarify the extent to which its operation must be conditioned by Article 21, thereby determining whether money laundering prosecutions will continue to rely on extended pre-trial custody as a systemic crutch, or be compelled to conform to a constitutional criminal process in which deprivation of liberty is strictly ancillary to, and not a substitute for, timely adjudication.
*Anil Tiwari, Partner
**Sanskruti Jinwal, Associate
[1] 2026: PHHC : 112452
[2] This structural anomaly concerning the schedule-link to predicate offences is explicitly identified and criticised in Nikesh Tarachand Shah v. Union of India, (2018) 11 SCC 1.
[3] (2018) 11 SCC 1
[4] 2022 INSC 757
[5] For the distinction between “appear” and “satisfied” as standards in bail provisions and its implications for presumption of innocence, see comparative analysis of TADA, PMLA and CrPC bail provisions in case law following Kartar Singh v. State of Punjab, (1994) 3 SCC 569 and Vijay Madanlal Choudhary v. Union of India, 2022 INSC 75.
[6] 2024 INSC 595
[7] Kartar Singh v. State of Punjab, (1994) 3 SCC 569
[8] 2024 INSC 739
[9] 2024 INSC 637
[10] 2026] 1 SCR 119
[11] 2026: PHHC : 112452
[12] Varun Puri v. Directorate of Enforcement, 2026: PHHC : 112452, introductory discussion drawing on Ulpian, Magna Carta, and Blackstone.