BNSS one year on: bail, custody, default release, trial in absentia, and the s.482 discretion

The Bharatiya Nagarik Suraksha Sanhita, 2023 — the BNSS — came into force on 1 July 2024, replacing the Code of Criminal Procedure, 1973. A year on, the practitioner-level architecture has substantially settled into shape. The text remained, on its face, broadly continuous with the CrPC — the structural choices about bail, custody, default release, inherent powers, and the recording of statements were largely preserved — but the textual changes that did occur have produced a substantial body of new doctrine, and a few have produced doctrinal confusion that the Supreme Court has been working to resolve.

This piece reads the year together. The substantive components — bail under s.480, default release under s.187, inherent powers under s.482, trial in absentia under s.356, the recording-of-statements architecture under s.183, and the protective regime under ss.367–369 — each have a developing doctrinal line that practitioners should understand. The point of the piece is to assemble them into a single architecture.

Bail conditions: Narayan v. State of MP and the trial-court reset

The most significant Supreme Court engagement of the year is the disposition of 22 April 2026 in Narayan v. State of Madhya Pradesh — reported as 2026 LiveLaw (SC) 426, decided in SLP (Criminal) No. 7011 of 2026 by a division bench of Justices J.K. Maheshwari and Atul S. Chandurkar. The bench addressed a question that had been recurring across trial courts since the BNSS came into force: whether the mandatory bail conditions under Section 480(3) — surrender of passport, weekly or daily reporting to a police station, restrictions on movement within districts or states — attach to non-bailable offences punishable with imprisonment up to seven years.

The text of Section 480(3) had supplied a clear answer: the mandatory conditions are stipulated for non-bailable offences punishable with imprisonment of seven years or more — that is, for the more serious end of the non-bailable cognisable spectrum. The condition architecture does not, on the text, apply to non-bailable offences punishable with imprisonment up to seven years.

Despite the textual clarity, trial courts across the country had been imposing the standard conditions in cases where they did not apply. The practical effect was substantial: a defendant whose case carried a maximum sentence of six years would, on grant of bail, find himself subject to passport surrender, weekly police-station attendance, and movement restrictions that he had no statutory basis to bear.

The Narayan disposition settled the question. The mandatory conditions under Section 480(3) attach to the seven-years-and-above category; they do not attach to the lower-than-seven-years category. The Court clarified that the conditions can be imposed only where the accused is charged with an offence "punishable with imprisonment which may extend to seven years or more or of an offence under Chapter VI, Chapter VII or Chapter XVII of the BNSS or abetment of, or conspiracy or attempt to commit, any such offence." The underlying case was a Madhya Pradesh Excise Act, 1915, prosecution carrying a maximum of three years' imprisonment; the Madhya Pradesh High Court had cancelled the appellant's bail; the Supreme Court restored it. Trial courts that have been imposing the Section 480(3) conditions in below-seven-years cases are, on the Narayan line, doing so without statutory authority.

The practical implication is that bail orders issued in cases below the seven-year threshold are open to challenge or modification where they include the mandatory conditions. The remedy is application to the trial court — or, where that fails, to the High Court — for variation of the bail conditions.

Default release under Section 187: the Kulkarni–Senthil Balaji split

The most doctrinally difficult question of the year — and the one that the Supreme Court has not yet definitively resolved — concerns the operation of Section 187(3) of the BNSS.

The provision, which is the successor to Section 167(2) of the CrPC, retains the default-release architecture: an accused whose investigation has not been completed within the prescribed period — 60 or 90 days depending on the maximum sentence — is entitled to release on bail if he is prepared and does furnish bail. The 60/90-day clock continues to operate as it did under the CrPC.

The doctrinal question Section 187(3) has produced concerns the police-custody architecture within the prescribed period. The text — which provides for police custody "in the whole or in part" within the 40 or 60-day window depending on the offence category — has been read as authorising the police to request police custody in fragments, rather than in a single 15-day continuous block.

The interpretation matters because the consequence of a fragmentary architecture is that an accused who has been released to judicial custody may be re-taken into police custody for a subsequent fragment within the window. The practical effect — which High Courts have identified as a harassment risk — is repeated re-taking, with the accused unable to settle into the judicial-custody framework and the magistrate facing repeated police-custody applications across the window.

Two competing High Court interpretations have emerged. The Kulkarni line — drawing on an older reading of the CrPC architecture — has read the BNSS provision conservatively, treating the police custody as a single 15-day block within the window and resisting the fragmentary reading. The Senthil Balaji line — taking the BNSS text at its more permissive face — has read the provision as authorising the fragmentary architecture, with the magistrate exercising discretion on each police-custody application.

As of the date of writing, no Supreme Court disposition under Article 141 has resolved the split. Different High Courts are operating on different interpretations; the magistrate's discretion within each High Court's territory is shaped by the prevailing line. The position is a doctrinal patchwork that practitioners must navigate by reference to the High Court line in the relevant jurisdiction.

The substantive defence available to an accused subject to a fragmentary police-custody application is the Kulkarni line where it applies, and an argument from the principle that magistrates should grant bail liberally — and resist re-taking into police custody — where the Senthil Balaji line is the local position. The principle has substantive content even within the more permissive reading.

The inherent powers under Section 482: the Chhattisgarh widening

Section 482 of the BNSS — the successor to Section 482 of the CrPC — preserves the inherent powers of the High Court to make orders necessary to give effect to the orders of the Code, to prevent abuse of the process of any court, or to secure the ends of justice. The text is, on its face, materially identical to the CrPC predecessor.

A textual change that the Chhattisgarh High Court has highlighted, however, concerns the absence of guiding factors that the BNSS text omits. The CrPC architecture had been read, across decades of authority, as containing certain implicit factors that guided the exercise of the inherent power — the rarest-of-rare formulation in quashing proceedings, the relationship to the substantive criminal-procedure framework, the principle that the power is to be sparingly exercised.

The Chhattisgarh High Court, in a disposition before Justice Goutam Bhaduri in the Parisha Trivedi matter, has read the BNSS text as deleting these guiding factors and as widening the discretion in consequence. The substantive holding — that the inherent power has, on the BNSS text, a less constrained operation than it did under the CrPC — is doctrinally significant, even as the wider impact is yet to be tested across other High Courts and, eventually, at the Supreme Court.

For practitioners, the operational consequence is that the Chhattisgarh line may produce, in the Parisha Trivedi tradition, a more readily-available quashing route for cases that — under the CrPC — would have been treated as too marginal to attract the inherent power. The strategic position should be calibrated to the High Court whose jurisdiction the matter falls under.

Trial in absentia under Section 356

The BNSS's introduction of a substantive trial-in-absentia architecture under Section 356 is one of its most consequential structural innovations. The provision permits trial in the absence of an accused who has been proclaimed an offender — and where the substantive prerequisites have been satisfied — including a 90-day wait period after proclamation, two warrants of arrest, paper publication in a local newspaper, and the appointment of state-funded counsel.

The framework has produced a developing body of doctrine on the substantive prerequisites — whether the 90-day clock runs from the date of the proclamation or from the date of its publication, what kind of paper publication satisfies the requirement, and the substantive role of state-funded counsel in protecting the absent accused's rights.

The direction of the doctrine has been protective. High Courts have been cautious in permitting trial in absentia where the prerequisites have not been properly satisfied, with paper publication and the appointment of competent counsel attracting the most scrutiny.

The recording-of-statements architecture under Section 183

Section 183 of the BNSS supplies the architecture for the recording of statements by witnesses before a magistrate. The provision preserves much of the CrPC s.164 architecture but introduces refinements that have produced their own developing line.

The Allahabad High Court disposition before Justice Saurabh Lavania — addressing the Kajal matter — has been the most significant engagement of the year. The holding addresses the discretion of the investigating officer to sponsor a witness for Section 183 recording, and the relationship of that discretion to a party application for the same recording.

The doctrinal frame is that the investigating officer's discretion is the substantive locus of the decision — not displaced by a party application requesting that the recording be done. The party application is, properly understood, a request to the investigating officer to exercise the discretion; it is not an independent route to the magistrate.

The protective regime under Sections 367–369

The BNSS's protective regime for accused who are unable, by reason of intellectual disability or otherwise, to make their defence — under Sections 367 to 369 — has produced one of the year's most consequential dispositions. The Kerala High Court, before Justice K. Babu, in the Thankappan matter, held that the protective regime applies to accused with conditions including Alzheimer's, and that the protection is retrospective in operation.

The retrospective reading is doctrinally consequential. The proposition is that an accused who, at the time of trial, falls within the protective regime is entitled to its operation regardless of when the substantive offence was alleged to have occurred. The protective doctrine attaches at the trial stage and looks at the accused's substantive ability to defend at that stage.

The architecture, drawn together

Read together, the year's jurisprudence supplies a working practitioner architecture.

On bail: Narayan settles the s.480(3) mandatory-condition question for the below-seven-years category. The trial-court reset is on its way.

On police custody: the Kulkarni–Senthil Balaji split persists, and the magistrate's discretion within each High Court's territory is shaped by the prevailing line. The doctrinal patchwork awaits Article 141 resolution.

On the inherent powers: the Chhattisgarh line — read in the Parisha Trivedi tradition — has widened the s.482 discretion, with the wider impact still to be tested.

On trial in absentia: the substantive prerequisites under s.356 are receiving close scrutiny.

On the recording of statements: the Kajal disposition treats the investigating officer's discretion as the substantive locus.

On the protective regime: Thankappan operates the s.367–369 architecture retrospectively for accused with intellectual disability or comparable conditions.

The architecture is substantive but incomplete. The Section 187(3) split is the most pressing open question, and the Supreme Court will, at some point, have to address it.

What the year has not produced

Three substantial gaps remain.

The first is a comprehensive Supreme Court engagement with the BNSS's evidentiary cross-references — particularly the operation of the BSA (Bharatiya Sakshya Adhiniyam) within the BNSS framework — which has been left to the High Courts.

The second is a definitive treatment of the relationship between the BNSS and the special legislation that continues to operate alongside it — the UAPA, the PMLA, the NDPS Act — where the BNSS's substantive bail and custody provisions interact with the special-statute architecture.

The third is a comprehensive treatment of the substantive role of state-funded counsel under the trial-in-absentia and protective-regime architectures, where the textual provisions have not been matched by an institutional framework for the recruitment, deployment, and oversight of such counsel.

These will need to be addressed.

Related editorial pieces

• BNSS Section 356 and trial in absentia: the proclaimed-offender architecture
• BNSS bail, one year on: practitioner notes
• Chhattisgarh High Court on Section 482 BNSS: the deletion of guiding factors and the widened discretion
• Kerala High Court, Thankappan, and the s.367–369 protective regime for accused with intellectual disability
• Allahabad High Court, Kajal, and the IO's discretion under s.183 BNSS
• BSA Section 63: the electronic-evidence architecture for the practitioner