For Reading a Rebuttable Presumption Against Custody Into § 204 CrPC

Recently, in Serious Fraud Investigation Office v. Aditya Sarda, the Supreme Court reversed the grant of anticipatory bail to the accused individuals who had not been arrested during investigation and had fully cooperated with the SFIO. Following the filing of a prosecution complaint under § 212(6) of the Companies Act, the Special Court issued non-bailable warrants at the very outset, without first summoning the accused. While no custodial interrogation had been sought earlier, the Supreme Court reversed this decision, holding that trial courts retain discretion under § 204 of the CrPC to issue either summons or warrant, if the gravity of the offence or considerations of public interest so require.

In this essay, I argue that where prosecutions follow non-custodial investigations, a constitutionally faithful interpretation of the CrPC demands courts to recognise a rebuttable presumption against post-charge sheet custody, especially in the absence of any demonstrated necessity for remand.

To begin with, a principled reading of the provision reveals that it is predicated on a constitutionally rooted, tiered approach to securing the accused’s presence. This is evident in its distinction between “summons-cases” and “warrant-cases,” where even within warrant cases, it vests the Magistrate with the discretion to issue either a summons or a warrant. Interestingly, it does not mandate the issuance of a warrant (even in warrant cases), far less a non-bailable one, as the initial mode of process.

§ 204 of the CrPC states:

“204. Issue of process.—(1) If in the opinion of a Magistrate taking cognizance of an offence there is sufficient ground for proceeding, and the case appears to be—
(a) a summons-case, he shall issue his summons for the attendance of the accused; or
(b) a warrant-case, he may issue a warrant, or, if he thinks fit, a summons, for causing the accused to be brought or to appear at a certain time before such Magistrate or (if he has no jurisdiction himself) some other Magistrate having jurisdiction…(emphasis mine)

This mode of understanding is perhaps best explained in the Supreme Court’s decision in Inder Mohan Goswami, where the Court held that non-bailable warrants should be issued only when summons or bailable warrants are unlikely to secure the accused’s presence. Such conditions may arise if the person is evading service, is likely to abscond, or poses a threat if left at large. The Court emphasised that summons or bailable warrants must be preferred wherever sufficient, and that any decision to issue non-bailable warrants must follow a careful scrutiny of facts and application of judicial mind, given the gravity of the consequences involved. 

Read in this light, summons come first, after which, if the accused fails to appear, bailable warrants follow. Only if the accused continues to evade process or demonstrates a risk of absconding, can non-bailable warrants then be issued. The logic behind this sequencing is fairly straightforward. Summons assume cooperation. Bailable warrants respond to hesitation or delay. While non-bailable warrants, then, address defiance.  

Against this backdrop, the decision in Aditya Sarda marks a troubling departure. By treating non-bailable warrants as an initial measure of compulsion (notwithstanding the absence of custodial investigation, evasion, or non-cooperation by the accused) the Court fails to appreciate the graded model embedded in § 204. It permits the most coercive form of process without first exhausting, or even attempting, the least restrictive.

Such a reading, I submit, destabilises both the clarity of Inder Mohan Goswami, and the broader constitutional structure that safeguards liberty through a graded model of restraint. When prosecutions arise from non‑custodial investigations and the accused have demonstrated full cooperation, trial courts must begin from a rebuttable presumption against post‑charge‑sheet custody, resorting to non‑bailable warrants only upon concrete proof of evasion, absconding risk, or public‑interest necessity. 

Anything less transforms § 204’s tiered model into a blunt instrument, eroding the liberty‑protective logic and unsettling the broader constitutional commitment that coercive process be both necessary and minimally impairing.

On the Supreme Court’s Asset Substitution Order in M3M — and Its Discontents

On 30.06.2025, the Supreme Court permitted M3M Group (M3M) the substitution of provisionally attached property under the Prevention of Money Laundering Act (PMLA) with alternate unencumbered assets (specifically — unsold commercial units from their flagship real-estate project). Interestingly, the Enforcement Directorate (ED) raised no objection. Instead, it acceded to the arrangement, subject to a series of nine conditions. These included the submission of a no-encumbrance certificate, an undertaking not to alienate the property, disclosure of funding sources, and full cooperation with ongoing investigation (a detailed enumeration of the conditions can be found here). Noting that the relief was confined to the facts of the case and not to be treated as precedent, the Court allowed the substitution to proceed.

In this essay, I pose a perhaps more philosophical concern that the Court’s substitution order in M3M raises. My only intention is for readers to reflect on a set of broader institutional questions: who, in practice, is actually in a position to benefit from this kind of judicial relief? Under what conditions does such a remedy become feasible? Is the availability of substitution shaped less by law and more by the capacity to access it?

Relief like this is not casually stumbled into. It likely requires seasoned legal teams, prior negotiations with the ED, and a capacity to marshal unencumbered assets at short notice. Is it fair to assume, then, that only certain kinds of accused with the backing of professionals and reputational capital can even attempt to propose such arrangements? Could it be that the very possibility of substitution is already filtered through resources and influence?

One might wonder how this plays out for the vast majority of individuals caught in the PMLA’s dragnet. When property is provisionally attached, is there usually any room for negotiation? Or does the process operate with a kind of automaticity? If so, does it mean that in select cases, this rigidity can be reinterpreted, massaged, or worked around? Is this simply good lawyering and responsiveness to context? Or does it risk creating two parallel enforcement experiences?

Is it possible that perhaps a certain posture helps too? That accused persons who arrive in court not just with clean assets but also with a narrative of cooperation are more likely to be accommodated? And if so, how many accused are in a position to even construct that narrative? Can individuals, with no resources to spare and no reputational leverage to lean on, realistically expect to be seen in the same light?

Should we then be asking whether discretion in anti-money laundering enforcement inevitably maps onto economic hierarchies? Or is that too cynical a reading of what courts try to do when confronted with complex, case-specific hardship? Perhaps the Court in M3M was simply responding to an unusual situation, with appropriate safeguards and no intention of setting precedent. But even if that being so, might its decision shape the expectations of similarly placed parties going forward? Could it quietly shift the boundary between what is impermissible and what is practically negotiable?

It is worth reflecting, too, on what this means for the ED. If substitutions become a viable option in certain cases, would that alter how the ED approaches attachment decisions? Would it create informal room for flexibility, for strategic consent, for behind-the-scenes negotiation? And if so, is that consistent with the PMLA’s stated architecture of rigidity and uniformity?

To conclude, the Supreme Court’s decision in M3M opens up more than just a procedural exception. By extending such relief only to those with resources to navigate its complexities, it casts a shadow over questions of equal access. These concerns, I submit, though wrapped in the caveat of non-precedential value, demand a closer attention.

Stretching the Exception — On Bihar Electoral Rolls Revision and the ECI's Overreach under Section 21 RPA

On 24 June 2025, the Election Commission of India (ECI) announced a Special Intensive Revision (SIR) of the electoral rolls in Bihar. It appears to be, on its face, an ambitious exercise. For over seven crore entries are to be verified, re-submitted, or deleted within a 90-day window. A first of its kind since 2003, the SIR aims to achieve three objectives: (a) ensure that every eligible citizen is enrolled and no one is excluded, (b) remove all ineligible voters from the rolls, and (c) eliminate dead, shifted, or absentee voters.

Coming just months before the state assembly elections, the ECI has justified this intensified exercise by pointing to factors such as rapid urbanisation, frequent migration, the influx of newly eligible young voters, underreporting of deaths, and the inclusion of names of foreign illegal immigrants. 

Writing on Constitutional Law and Philosophy, Vasudev Devadasan has already outlined the gravity of the moment. In his tightly argued breakdown, he highlights: (i) the SIR exercise risks disenfranchising existing voters, particularly those added after the 2003 revision, by imposing retrospective verification burdens with no statutory basis, (ii) the ECI’s documentary exclusions (notably Aadhaar, EPIC voter IDs, and ration cards) are misaligned with both statutory mandate and administrative reality, and (iii) the compressed timeline, wedged between a summer announcement and an autumn election, compromises the ability of voters to meaningfully respond to errors or exclusions.

In this essay, I do not intend to rehearse those arguments. Instead, I aim to highlight a concern that is less visible, but perhaps equally structural.

I submit that the ECI’s discretion under Section 21(3) of the Representation of the People Act (RPA) has been misread to authorise a form of jurisdictional expansion it was never meant to enable.

Section 21(3) of the RPA reads as follows:

“Notwithstanding anything contained in sub-section (2), the Election Commission may at any time, for reasons to be recorded, direct a special revision of the electoral roll for any constituency or part of a constituency in such manner as it may think fit:

Provided that subject to the other provisions of this Act, the electoral roll for the constituency, as in force at the time of the issue of any such direction, shall continue to be in force until the completion of the special revision so directed.”

This clause is the statutory basis on which the ECI has grounded its SIR of the Bihar electoral rolls. On a plain reading, the text suggests a wide scope. It allows the Commission (I use ECI and “the Commission” interchangeably) to order a “special revision” at “any time,” and to do so in a manner it “thinks fit.” However, statutory breadth is never a free pass. 

To begin with, Section 21(3) is an exception to the general rule under Section 21(2), which contemplates annual revisions of the electoral roll. Seen in this light, the statutory scheme becomes relatively clear — for it indicates that routine revisions are to form the foundation of electoral roll maintenance, while special revisions are meant to step in only when routine mechanisms fall short. So understood, Section 21(3) functions as a safeguard, not an alternative pathway. The power to conduct special revisions must therefore remain tethered to the structure within which it is situated.

In Bihar, however, the ECI has used Section 21(3) to launch what is effectively a ground-up reinvention of the electoral roll. Over seven crore entries are to be reverified, updated, or deleted within a 90-day window. This is no minor intervention. It is an exercise of staggering scale, with deep electoral consequences. The criteria invoked (migration, urbanisation, underreporting of deaths, and presence of alleged illegal immigrants) are undeniably serious, but none of them justify an inversion of the presumption of continuity that the RPA’s electoral roll provisions rest upon. 

This suggests a deeper structural problem. The Commission appears to be stretching the provision’s scope beyond what the Parliament had likely contemplated. When a statute authorises a “special revision,” it assumes that the normal revision has occurred or will occur in accordance with law, and that the special revision will operate as a limited corrective mechanism. The “reasons to be recorded” requirement perhaps even reinforces this logic. On such a reading, special revisions, in other words, would respond to identifiable contingencies, and not broad governance concerns. To allow otherwise risks blurring the line between the Commission’s role as a supervisor and the operational logic of an executive bureaucracy.

This becomes clearer through illustration. For instance, take “migration.” Internal migration in India, especially in states like Bihar, is a continuous phenomenon. Treating this as a trigger for a state-wide intensive revision has potential to turn what should be an exceptional measure into a routine one. Similarly, while underreporting of deaths is a long-standing challenge, it has never, until now, been seen as grounds for a statewide electoral overhaul.

Documentation

This page curates a list of documents and judgements from some of the more consequential cases argued before the Supreme Court in the past year or so. 

Constitutionality of the Electoral Bond Scheme

Judgement declaring the scheme unconstitutional (here)

Chandrachud CJ.'s majority opinion (here)

Khanna J.'s majority opinion (here)

Constitutional Validity of Section 6A of the Citizenship Act

Judgement upholding validity (here)

Kant J.'s majority opinion (here)

Chandrachud CJ.'s concurring opinion (here)

Pardiwala J.'s dissenting opinion (here)

Aligarh Muslim University's Minority Status

Judgement overruling Azeez Basha (here)

Chandrachud CJ.'s majority opinion (here)

Kant J.'s dissenting opinion (here)

Datta J.'s dissenting opinion (here)

[A quick note — I have a piece up on NLSIR Online where I engage with Datta J.’s dissent and outline some of the concerns that it raises. For those interested, the link is here.]

Sharma J's dissenting opinion (here)

Pendency of Bills before Tamil Nadu Governor 

Judgement reaffirming constitutional limits on gubernatorial inaction (here)

Supreme Court records data submitted by Tamil Nadu Governor on pending Bills (here)

About

This blog is a modest effort to write more regularly. Much of what I work on never finds its way - held back by doubts about novelty, structure, or fit. 

In setting up this space, I hope to overcome that inertia. 

Inspired in no small part by Mr. Abhinav Sekhri’s The Proof of Guilt blog, this platform will carry reflections on constitutional questions, commentary on litigation involving economic offences, inquiries into criminal procedure, and occasional forays into public law more broadly. 

The scope is intentionally wide, shaped less by subject boundaries and more by curiosity and continuity.