Part 2 | Andrabi engages with the Union’s deferring tactic in UAPA bail cases

This commentary is divided into two parts (Read Part 1 here). The first part examines the unprecedented reliance on data in Andrabi  as an argument against pre-trial detention in bail jurisprudence under special statutes. 

On 18 May 2026, a Division Bench of Justices B.V. Nagarathna and Ujjal Bhuyan, in Syed Iftikar Andrabi v NIA (Andrabi) granted bail to an undertrial prisoner—Syed—who was incarcerated for over five years and eleven months on charges of narco-terrorism under the Unlawful Activities (Prevention) Act, 1967 (“UAPA”) and the Narcotic Drugs and Psychotropic Substances Act, 1985 (“NDPS Act”). In the process, the Bench criticised two rulings—Gulfisha Fatima v State (2026) (“Gulfisha”) and Gurwinder Singh v State of Punjab (2024) (“Gurwinder”) for hollowing out a three-judge bench decision, Union of India v. K.A. Najeeb (2021) (“Najeeb”). Najeeb held that statutory conditions of bail under the UAPA would “melt down” in light of prolonged incarceration. Gulfisha, which denied bail to activist Umar Khalid, and Gurwinder were authored by Justice Aravind Kumar.

Five days later, on 22 May, a different Division Bench of Justices Kumar and P.B. Varale in Tasleem Ahmed v State (NCT of Delhi) (“Tasleem”) referred the question of how Article 21 of the Constitution is to be applied against the statutory bar on bail under Section 43-D(5) of the UAPA to a larger bench. This was done at the request of the Union government, with the Additional Solicitor General S.V. Raju pointing to the Andrabi decision.

Both decisions flow from the same upstream fault line in the Court’s jurisprudence. In Andrabi, the Court criticised Gulfisha and Gurwinder. In Tasleem, the Court, rather than add a further competing formulation, referred to the conflict for “authoritative resolution.”

This commentary examines Andrabi’s rejection of what has become a standard State offer: delaying the grant of bail by one more year. Lastly, it examines, most importantly, the question of whether the reference itself was necessary, or whether it was optional? 

“Not now, Try after a year”: A Distinct Constitutional Right

In Andrabi, the Court rejected the State’s submission for delaying the consideration of bail “after one year” on the ground that the trial would be completed by then. The rejection rests on two decisions—High Court Bar Association, Allahabad v State of Uttar Pradesh (2024) (“HCBA”), and the subsequent Division Bench decision in Rup Bahadur Magar v State of West Bengal (2024). 

The “Not now, Try after one year” as a deferral mechanism evolved as a response to the post-Najeeb landscape. Once the Court recognised that prolonged incarceration and delayed trial could override Section 43-D(5), the State and, at times, the Courts developed a counterweight. Rather than grant bail on account of delay, the court would direct the trial court to conclude proceedings within a fixed period. The accused would remain in custody; the trial court would receive a deadline it could rarely honour. When the deadline passed without compliance, the accused would return to seek what should have been granted at the outset.

In Harpreet Singh Talwar, the Court declined bail while granting liberty to reapply after six months. Similarly, in Gulfisha, the Court rejected Umar Khalid and Sharjeel Imam’s bail but granted liberty to approach for bail after 1 year of the Order (i.e., to say 6+ years incarceration is sufficient for bail but not 5+ years). These orders are not isolated orders; they reflect a systematic practice of acknowledging the constitutional problem while postponing the constitutional remedy.

In HCBA, the 5-Judge bench addressed the issue of whether the Supreme Court could, under Article 142, direct automatic vacation of High Court interim stays and order time-bound disposal of stayed proceedings. The Court held that constitutional courts should not ordinarily direct the disposal of cases before trial courts within a time span. Three reasons were stated: (a) trial courts carry heavy arrears and cannot prioritise out-of-turn cases without displacing older matters where accused persons have been in custody even longer; (b) constitutional court directions create inequitable advantages for litigants who can afford to approach these courts; and (c) superior courts should not interfere with the day-to-day functioning of subordinate courts by specifying out-of-turn disposal timelines. 

Andrabi imports these reasons directly into the bail context through Rup Bahadur Magar, which had held that High Courts passing time-bound trial disposal orders while rejecting bail applications were acting contrary to the law laid down in HCBA. The invocation of HCBA in Andrabi in response to the State’s submission, therefore, does more than cite a precedent; it establishes that the “Not now, Try after one year” offer is impermissible as a response to a bail application.

Two Distinct Rights, Not One.

The more foundational point in Andrabi’s treatment of this issue is that the right to bail and the right to a speedy trial are distinct constitutional rights. While they flow from Article 21, they address different concerns and command different remedies. 

The right to a speedy trial protects the accused’s interest in having the question of guilt or innocence adjudicated without unreasonable delay. It is primarily a right against the State’s failure to conduct proceedings expeditiously. Its remedy is procedural acceleration, directions to the trial court to proceed, and, ultimately, consequences for persistent non-compliance. 

On the other hand, the right not to be subjected to pre-trial detention as a form of pre-adjudication punishment is distinct. The accused is presumed innocent. Detention prior to conviction is a prima facie interference with the right to liberty under Article 21. Najeeb holds that when this interference becomes “unduly prolonged,” and the trial is “unlikely to conclude within a reasonable time,” the constitutional court is “obligated to enlarge” the accused on bail, not because the trial will be faster, but because continued detention has become impermissible. 

When a court delays the grant of liberty with a direction to conclude the trial, it has answered the wrong question. The accused is not seeking a faster adjudication in the abstract. They are seeking not to be detained without conviction. A direction that the State provide faster proof of guilt or innocence does not address the present deprivation of liberty. It merely promises that the deprivation will end sooner if the trial court complies, which, as HCBA recognised, it structurally cannot in most cases. 

Andrabi is explicit that this approach is improper. The “Not now, Try after one year” direction had, until Andrabi, survived without formal doctrinal rejection. Andrabi does not merely criticise the practice; it holds that courts must not resort to it. Trial courts and High Courts receiving bail applications under the UAPA are not permitted to treat the “expedite the trial” direction as a substitute.

Was the reference necessary? 

The law on when a smaller bench may refer a matter for reconsideration by a larger bench is, on its face, well settled. In Central Board of Dawoodi Bohra Community v. State of Maharashtra (2004), the Court held that a larger bench binds all smaller and co-equal benches. A bench of a lesser quorum cannot doubt a larger bench’s ruling; it may only request the Chief Justice to constitute a larger bench. Only a bench of co-equal strength may express doubt about an earlier co-equal bench’s decision, whereupon the matter goes to a bench of larger quorum for resolution.

The situation before the Tasleem was the following. Najeeb was a 3-Judge Bench decision. Gurwinder and Gulfisha were 2-Judge Bench decisions that had departed from Najeeb’s ratio, according to Andrabi. Andrabi itself was a 2-Judge Bench decision. Under the Dawoodi Bohra framework, a two-Judge Bench that disagrees with a two-Judge Bench decision cannot, in the ordinary course, do more than express a doubt and refer the matter to a three-Judge Bench. 

The situation in Tasleem was not one in which the bench was doubting a coordinated two-Judge bench decision. The conflict, as framed by the bench, was between Najeeb (3-Judge bench) and the subsequent 2-Judge bench decisions that had allegedly misapplied it.

This distinction is important. Dawoodi Bohra specifically addresses the situation where a bench of a lesser quorum doubts the correctness of a decision by a bench of a larger quorum. What it does not address is where a 2-judge bench doubts the application of a larger bench decision by a co-equal bench.  

In Andrabi, the bench had addressed the pure two-Judge-doubting-two-Judge scenario when it held that Gurwinder and Gulfisha could not be followed as a matter of precedent because they were inconsistent with Najeeb. In Tasleem, the bench characterised the situation as a “perceived divergence” between coordinate benches on the application of a Najeeb, and referred. It correctly observed that a coordinate bench cannot, by “language of reservation,” achieve what it cannot achieve by declaration of law. A doubt expressed by Andrabi is not a declaration. If the earlier view is thought to be inconsistent with a larger bench decision, the proper answer is “authoritative resolution”. That reasoning is internally consistent and sound. 

What the Tasleem bench does not address, however, is whether Andrabi itself had already provided the authoritative resolution. Andrabi held that Najeeb must be followed. That is not a doubt; it is a determination. 

The conflict between Andrabi (which says Najeeb governs) and Gulfisha (which says Najeeb is a narrow exception) is a conflict between coordinate benches of equal strength. And a coordinate bench cannot simply declare the earlier coordinate bench wrong; it must request placement before a larger bench. On this reading, Tasleem’s reference was not merely institutionally preferable, but was required.

Time to resolve Article 21’s (divided) promise? 

Andrabi will be remembered primarily as a stare decisis corrective. Tasleem, in turn, will be seen as not merely appropriate but a required step to resolve the conflict between coordinate two-Judge bench decisions on the application of a three-Judge bench precedent. The larger bench that will be constituted (if and when it is) has, therefore, an unusual combination of responsibilities. It must resolve the precedential conflict, which is, by now, not especially difficult given the unambiguous foundation of Najeeb. But it must also address the questions that Najeeb itself left open and that the five-year jurisprudential sequence has made urgent: how long is too long, what factors may and may not defeat the constitutional inquiry, whether conviction rate data can ground proportionality challenges to statutory restrictions, and whether courts may prospectively suspend the right to seek bail as a condition of granting it to co-accused. 

Whether the resolution the larger bench provides will be adequate to the constitutional problem that UAPA prosecutions have produced is a question that will be answered in the implementation. The law, when finally settled, will be only as effective as those who invoke it, and only as just as those who apply it.