Arresting the Arbitrariness: A Critical Examination of Pankaj Bansal v. Union of India

The Supreme Court’s judgment in Pankaj Bansal v. Union of India (2023INSC866) marks a decisive moment in India’s evolving jurisprudence on the prevention of money laundering. Confronted with a pattern of conduct by the Directorate of Enforcement that the Bench characterised as manifesting a “complete and utter lack of bonafides,” the Court laid down two propositions of lasting constitutional significance: first, that mere oral communication of grounds of arrest cannot satisfy the mandate of Article 22(1) of the Constitution read with Section 19(1) of the Prevention of Money Laundering Act, 2002, and that a written copy of such grounds must henceforth be furnished to every person arrested under the Act as a matter of course; second, that a remand order — however formally obtained — cannot cure or validate an arrest that was unlawful in its inception. This article critically examines both propositions, the factual matrix that produced them, and the Court’s broader censure of the ED’s conduct.

I. Introduction

The Prevention of Money Laundering Act, 2002 (“the Act” or “PMLA”) is, by design, a statute of exceptional severity. Section 45 imposes onerous twin conditions on the grant of bail; Section 50 enables compulsory examination of witnesses without the privilege against self-incrimination that ordinarily attends criminal interrogation; and Section 19 vests a power of arrest in officers of the Directorate of Enforcement (“ED”) on the basis of their subjective satisfaction, without the pre-arrest scrutiny of a magistrate that characterises ordinary arrests under the Code of Criminal Procedure, 1973.

The constitutional validity of these provisions was confirmed by a three-Judge Bench of the Supreme Court in Vijay Madanlal Choudhary v. Union of India, 2022 (10) SCALE 577, which held — inter alia — that the stringent safeguards built into Section 19 themselves justified the severity of the twin conditions for bail under Section 45. The corollary of that holding is clear: if the safeguards under Section 19 are routinely diluted or circumvented in practice, the very constitutional foundation on which Vijay Madanlal Choudhary rests is undermined.

It is precisely this corollary that Pankaj Bansal v. Union of India brings into sharp relief. The case is not merely about the alleged procedural lapses in two individual arrests; it is, at a deeper level, about whether the ED’s exercise of its extraordinary powers under the PMLA is subject to rigorous accountability — and whether the courts are willing to enforce that accountability.

II. Factual and Procedural Background

The origins of these proceedings lie in FIR No. 0006 of 17.04.2023 registered by the Anti-Corruption Bureau, Panchkula, Haryana, under Sections 7, 8, 11, and 13 of the Prevention of Corruption Act, 1988 read with Section 120B IPC. The FIR named a sitting Special Judge (who presided over CBI and ED cases), his nephew (who served as a legal officer in the M3M Group), and Roop Bansal, a promoter of the M3M Group, as accused. The allegation was that Sudhir Parmar, the Special Judge, had shown favouritism to litigants before him, including the owners of M3M Group.

Separately, between 2018 and 2020, a total of 30 FIRs had been registered against the IREO Group by its allottees. On the basis of those FIRs, the ED had recorded Enforcement Case Information Report No. GNZO/10/2021 dated 15.06.2021 — referred to in the judgment as the “first ECIR” — targeting the IREO Group and its Vice-Chairman, Lalit Goyal. Neither the M3M Group nor the appellants, Pankaj Bansal (son) and Basant Bansal (father), were named in the first ECIR or in any of the constituent FIRs.

After the ED arrested Roop Bansal in connection with the first ECIR on 08.06.2023, both Pankaj Bansal and Basant Bansal apprehended that action would be taken against them as well. They accordingly secured interim protection by way of anticipatory bail from the Delhi High Court on 09.06.2023, valid until 05.07.2023. The ED immediately filed Special Leave Petitions before the Supreme Court against these bail orders.

Having recorded the second ECIR on the evening of 13.06.2023, the ED issued summons to the appellants at 06.15 pm the same day, requiring their appearance at 11.00 am on 14.06.2023 in connection with the first ECIR — in which they had interim protection. When the appellants appeared in compliance with those summons, Pankaj Bansal was served with fresh summons at 04.52 pm on 14.06.2023, requiring his appearance before a different Investigating Officer at 05.00 pm the same day in connection with the second ECIR. Basant Bansal was arrested at 06.00 pm and Pankaj Bansal at 10.30 pm on 14.06.2023.

The arrested persons were taken to Panchkula, where the Vacation Judge/Additional Sessions Judge passed an order dated 15.06.2023 remanding them to ED custody for five days, without — as the Supreme Court later found — any application of judicial mind to whether Section 19 of the PMLA had been complied with. Subsequent remand orders extended their custody further. Their writ petitions before the Punjab & Haryana High Court were dismissed in limine by a Division Bench on 20.07.2023 and 26.07.2023, prompting the appeals to the Supreme Court.

III. Issues Examined by the Supreme Court

The Bench, speaking through Sanjay Kumar, J., identified and resolved the following issues:

IV. The Legal Framework: Section 19 PMLA and Article 22(1)

A. Section 19 PMLA — Conditions Precedent for Arrest

Section 19(1) of the PMLA authorises specified officers of the ED to arrest a person if they have “reason to believe” — based on “material in their possession” — that such person is guilty of an offence punishable under Section 4 of the Act. Three procedural requirements attach to this power: (i) the reasons for the belief must be recorded in writing; (ii) the arrested person must be “informed of the grounds for such arrest”; and (iii) a copy of the arrest order, together with the material in the officer’s possession, must be forwarded to the Adjudicating Authority in a sealed envelope under Section 19(2). Under Section 19(3), the arrested person must be produced before a Special Court or Magistrate within 24 hours.

B. Article 22(1) of the Constitution

Article 22(1) provides that no person who is arrested shall be detained in custody without being informed, “as soon as may be,” of the grounds for such arrest. This is a fundamental right of the arrested person, directly enforceable and not susceptible to legislative abridgement except in the circumstances specified in Article 22(3).

C. The Framework Set by Vijay Madanlal Choudhary

In Vijay Madanlal Choudhary, the Supreme Court held that Section 19 provides “inbuilt safeguards” that are “equally stringent and of higher standard” compared to analogous provisions in the Customs Act, 1962. The Bench further held that informing the arrested person of the grounds of arrest satisfies the mandate of Article 22(1), and that it is not mandatory to supply a copy of the ECIR, which may contain sensitive investigative material. What remained unresolved in Vijay Madanlal Choudhary — and what Pankaj Bansal directly addresses — was the precise mode of communication of grounds of arrest: is verbal communication sufficient, or must a written copy be supplied?

V. The Court’s Analysis — Issue by Issue

A. The Conduct of the ED: Malice in Law

Before examining the technical compliance issues, the Court undertook a searching review of the circumstances in which the second ECIR was recorded and the arrests were made. The Bench found the chronology to be “a rather poor reflection” on the ED’s “style of functioning.” Three specific findings are notable:

First, the second ECIR was based on a FIR that had been registered on 17.04.2023 — nearly two months before the ECIR was recorded on 13.06.2023. The ECIR was recorded only four days after the appellants obtained anticipatory bail from the Delhi High Court. The Court found it inconceivable that the ED was unaware of the FIR when the appellants were before the High Court, yet it was never disclosed to the High Court.

Second, the ED arrested the appellants within 24 hours of recording the ECIR, purportedly after “preliminary investigations” and on the basis of a formed opinion that they were guilty. The Court found it difficult to accept that the Investigating Officer could have formed such a clear opinion within so short a period.

Third, several factual assertions made by the ED in its reply affidavit — including that Pankaj Bansal had been summoned multiple times and had failed to comply, and that his name appeared in the FIR — were found by the Court to be factually incorrect.

“The way in which the ED recorded the second ECIR immediately after the appellants secured anticipatory bail in relation to the first ECIR, though the foundational FIR dated back to 17.04.2023, and then went about summoning them on one pretext and arresting them on another, within a short span of 24 hours or so, manifests complete and utter lack of bonafides.” — Para 22, Pankaj Bansal v. Union of India (2023INSC866)

Drawing on State of Punjab v. Gurdial Singh, (1980) 2 SCC 471, Collector v. Raja Ram Jaiswal, (1985) 3 SCC 1, and Ravi Yashwant Bhoir v. Collector, (2012) 4 SCC 407, the Court applied the doctrine of malice in law. The essential test — stated with precision in Gurdial Singh — is whether the power was exercised for ends beyond the sanctioned purposes of the statute, goaded by extraneous considerations. The Court’s finding, in substance, was that the prompt retaliatory recording of the second ECIR to circumvent the anticipatory bail already granted was a classic instance of fraud on power, irrespective of whether moral turpitude could be attributed to individual officers.

B. The Remand Order — Duty of the Magistrate under Section 167 Cr.P.C.

The Court reaffirmed and elaborated the duty of the remand court when an arrest is made under Section 19(1) PMLA. Drawing on V. Senthil Balaji v. State (Criminal Appeal Nos. 2284–2285 of 2023, decided on 07.08.2023), the Court held that the Magistrate or Special Court acting under Section 167 Cr.P.C. is “the appropriate authority who has to be satisfied about the compliance with safeguards as mandated under Section 19.” Non-compliance with Section 19 must benefit the arrested person and entitles him to release.

Examining the remand order of 15.06.2023, the Court found it to be a paradigm of judicial abdication. The Vacation Judge had not recorded any finding that he had perused the grounds of arrest or verified that the ED had recorded reasons to believe the appellants were guilty. The order merely noted the “seriousness of the offences” and the “stage of investigation” and remanded the appellants to ED custody. The sentence “all the necessary mandates of law have been complied with” appeared — by virtue of the word “further” — to be a continuation of the prosecution’s case, not the court’s own finding.

The Court additionally affirmed the three-Judge Bench decision in Madhu Limaye v. State, (1969) 1 SCC 292, which held that an order of remand would not cure constitutional infirmities in the underlying arrest if it was passed without proper application of judicial mind. The Court thus rejected the ED’s reliance on cases holding that the legality of arrest becomes irrelevant once a remand order is passed — rightly distinguishing those cases as being confined to the context of habeas corpus petitions where the legality of the arrest as at the date of the return of the rule was the only question.

C. The Central Question: Must Grounds of Arrest be Furnished in Writing?

This is the doctrinal centrepiece of the judgment and the point on which the Court broke new ground. The Section 19(1) requirement is that the authorized officer shall “inform” the arrested person “of the grounds for such arrest.” The word “inform” is not defined. In practice, the ED had adopted divergent approaches across the country: in some jurisdictions, a written copy of the grounds of arrest was furnished to the arrested person; in others, the grounds were merely read out, or the arrested person was permitted to read them.

The ED relied on two High Court precedents supporting oral communication as sufficient: the Delhi High Court in Moin Akhtar Qureshi v. Union of India (WP(Crl.) No. 2465 of 2017, decided on 01.12.2017), which held that Section 19 uses “informed” rather than “communicate” and therefore does not require written supply; and the Bombay High Court in Chhagan Chandrakant Bhujbal v. Union of India, 2017(1) AIR Bom R (Cri) 929, which held that there is “no statutory requirement” of communication in writing.

The Court rejected both these precedents and the reasoning underlying them, for the following reasons:

The purposive reading of Article 22(1): The mode of communicating the grounds of arrest must be “meaningful so as to serve the intended purpose.” The purpose is twofold: first, to enable the arrested person to understand why he is being held; second, to equip him to seek legal counsel and satisfy the stringent twin conditions under Section 45 PMLA to obtain bail. Unless the arrested person has knowledge of the specific grounds on which the authorized officer arrested him, he cannot effectively demonstrate before the Special Court that there are reasonable grounds to believe he is not guilty — which is the first of the twin conditions for bail. Oral reading of a potentially voluminous document, to a person in the shock and distress of arrest, cannot serve this constitutional purpose.

The Rule 6 argument: Rule 6 of the Prevention of Money Laundering (The Forms and the Manner of Forwarding a Copy of Order of Arrest) Rules, 2005, prescribes a standard Arrest Order in Form III which is used uniformly across the country. Yet the same form produces divergent practices — written supply in some jurisdictions, oral reading in others. This “dual and disparate procedure” on the basis of the very same prescribed format cannot be legally countenanced.

The evidentiary argument: Where grounds are merely read out and not supplied in writing, any dispute as to compliance reduces to the word of the arresting officer against the word of the arrested person. As the facts of Basant Bansal’s case illustrated — where he refused to sign the grounds, the ED produced witnesses to certify oral reading — this is a deeply unsatisfactory state of affairs that is entirely avoidable.

The redaction solution: The Court addressed the concern — expressed in Vijay Madanlal Choudhary — that sensitive investigative material must be protected. It held that while it is not necessary to supply all material forwarded to the Adjudicating Authority under Section 19(2), the grounds of arrest themselves are personal to the arrested person and ordinarily contain no sensitive third-party material. If sensitive material is included in the grounds, the authorized officer may redact it and furnish an edited copy — this is not an obstacle to written communication.

VI. The Law Points Laid Down — A Structured Summary

VII. Critical Evaluation

A. On the Written-Grounds Requirement

The Court’s holding on the written-grounds requirement is, with respect, both correct and constitutionally necessary. The argument that “informed” in Section 19(1) requires less than written supply is a textually thin one and is plainly inconsistent with the purpose that the provision is meant to serve. The PMLA is a statute under which bail is extraordinarily difficult to obtain. The twin conditions under Section 45 require, in substance, that the accused demonstrate to the Special Court that there are reasonable grounds to believe he is not guilty — a formidable burden. An accused who does not have a written record of the grounds on which the ED arrested him, and who may have heard those grounds read out to him at a moment of extreme stress, is effectively disabled from meeting this burden.

The Court’s purposive reasoning — linking the Article 22(1) right to the bail framework under Section 45 — is particularly important. It correctly identifies the structural relationship between the procedural safeguards at arrest and the substantive rights in subsequent proceedings. One may, however, observe that this reasoning was already implicit in Vijay Madanlal Choudhary and could have been extracted from that judgment earlier.

B. On the Validity of Remand and its Curative Effect

The Court’s reaffirmation that a remand order does not cure an unlawful arrest is constitutionally impeccable and doctrinally essential. A different rule would render the safeguards in Section 19 entirely nugatory: an arresting officer could bypass all procedural requirements confident that the subsequent remand order would legitimise the custody. It would also convert Section 167 Cr.P.C. proceedings into a rubber-stamp exercise — as, indeed, the Panchkula remand order on the facts appeared to be.

The Court’s express finding that the Vacation Judge’s remand order of 15.06.2023 reflected “total failure on his part in discharging his duty” is an unusually frank judicial criticism of a subordinate court. Its value lies not only in the resolution of the immediate appeal but as a reminder to remand courts across the country that they are not passive conduits for the executive’s detention decisions. The verification of compliance with Section 19 PMLA is a judicial function, not an administrative one.

C. On the Doctrine of Malice in Law

The Court’s application of the malice-in-law doctrine to the ED’s conduct is analytically sound and bears emphasis. The doctrine, as crystallised in Gurdial Singh, does not require proof of moral turpitude or personal animus. It is satisfied where the power is exercised for ends different from those for which it was conferred, or where extraneous or irrelevant considerations enter the exercise. On the facts of this case — the deliberate suppression of the second FIR before the Delhi High Court, the overnight recording of the second ECIR after anticipatory bail was granted, the arrests within 24 hours on the basis of purported “preliminary investigation” — it is difficult to disagree with the Court’s conclusion.

What is perhaps less satisfactorily resolved is the institutional consequence of this finding. The Court sets aside the arrests and orders of remand, and orders the release of the appellants. But it does not direct any inquiry into the conduct of the officers concerned, nor does it invoke the contempt jurisdiction or Section 62 PMLA (non-compliance with Section 19). The inference is that the Court considered the setting aside of the arrests to be a sufficient institutional response. Whether that is adequate to deter repetition of such conduct is a separate question.

D. On Non-Cooperation as a Ground for Arrest

Paragraph 25 contains an important holding that has received insufficient attention in commentary. The Court held that a person summoned under Section 50 PMLA cannot be arrested under Section 19 merely because he has been “evasive” or non-cooperative in responding to questions. This is constitutionally correct: Article 20(3) protects against compelled self-incrimination, and a summons under Section 50 does not convert the summoned person into someone who is obliged to incriminate himself. The fact that a person exercises caution in answering questions — whether because he is aware of his constitutional rights or otherwise — cannot, without more, be treated as evidence of guilt sufficient to sustain an arrest under Section 19.

This holding is important because there is an evident structural temptation for the ED to use Section 50 summons as a mechanism to compel answers, and then to treat any claimed evasiveness as a basis for Section 19 arrest. The Court’s ruling places a clear limit on this approach.

E. Observations on Institutional Accountability

One aspect of the judgment that deserves broader attention is the Court’s characterisation of the ED as a “premier investigating agency” that is “not expected to be vindictive in its conduct” and must be seen to be “acting with utmost probity and with the highest degree of dispassion and fairness.” This is not merely rhetorical. It reflects a constitutional expectation that agencies vested with extraordinary powers under exceptional legislation will exercise those powers with correspondingly heightened fidelity to the rule of law.

The facts of Pankaj Bansal — the suppression of the foundational FIR before the bail court, the retaliatory recording of the ECIR, the factually incorrect affidavit before the Supreme Court — suggest that this expectation was far from met in the case at hand. The Court’s response was proportionate: it set aside every consequential order from the arrest through to the remand, and directed immediate release. The deterrent effect of this outcome on future conduct by the ED remains to be seen.

VIII. Impact and Subsequent Developments

The judgment in Pankaj Bansal had an immediate and direct impact on PMLA jurisprudence. The written-grounds requirement has since been applied and reinforced by subsequent benches. The Supreme Court’s later decision in Prabir Purkayastha v. State (NCT of Delhi), (2024) 8 SCC 254, affirmed that a violation of Article 22(1) vitiates the arrest and the initial remand, and that the filing of a charge sheet does not cure such a violation. The Court in Vihaan Kumar v. State of Haryana, 2025 (5) SCC 799, reiterated non-compliance with the written-grounds requirement as vitiation of the arrest. Most recently, Mihir Rajesh Shah v. State of Maharashtra, (2026) 1 SCC 500, consolidated the position further, requiring that the grounds be in writing, in a language understood by the arrestee, and furnished at least two hours before the remand hearing.

The rule in Pankaj Bansal may therefore be seen as the foundational moment in a developing line of constitutional criminal procedure that progressively tightens the accountability standards for PMLA arrests. The central insight — that the right to be informed of the grounds of arrest must be made effective, not merely formal — has proven to have considerable doctrinal fertility.

IX. Conclusions

Pankaj Bansal v. Union of India (2023INSC866) is a judgment of considerable constitutional importance and, it is submitted, of lasting doctrinal value. The following principal conclusions emerge:

What makes this judgment enduringly significant is the Court’s recognition that the constitutional validity of the PMLA’s extraordinary provisions — confirmed in Vijay Madanlal Choudhary — was premised on the effectiveness of the safeguards in Section 19. If those safeguards are reduced to formal incantations, the constitutional foundation of the statute is eroded from within. Pankaj Bansal reasserts, with clarity and conviction, that constitutional safeguards must be real, not merely nominal — and that courts will not be passive spectators when executive agencies treat constitutional rights as administrative inconveniences.

⊕ Table of Cases Referenced