Advitiya Vaibhav Rai *
Introduction
Multiple individuals accused under theUAPA (The Unlawful Activites (Prevention) Act, 1967) have been in prolonged pre-trial incarceration. Their cases are not anomalies. They are foreseeable consequences of a bail regime that was designed through statute and successive judicial interpretation, to make bail nearly impossible. When the Supreme Court in Syed Iftikhar Andrabi v. National Investigation Agency, Jammu held in May 2026 that bail remains the rule even under UAPA and that prolonged incarceration without trial is a ground for release under Article 21, it offered what looked like a correction. It also revealed something troubling: that coordinate benches of the same Court had spent years quietly dismantling the only constitutional protection standing between a UAPA accused and indefinite detention. This piece argues that the larger bench reference triggered by Andrabi is necessary, but will remain insufficient unless the Court is willing to confront the structural questions that produced this crisis in the first place.
The Structural Flaws in the Legislation
Section 43D (5) of the UAPA states that no person accused of offences under Chapters IV or VI shall be released on bail if on the perusal of the prosecution’s case diary or chargesheet, the court finds the accusations prima facie true. This prima facie threshold has not been defined in the statute or by the courts in any judicial pronouncement. This has resulted in the courts assessing the case solely on the basis of prosecution’s own documents which results in a structurally circular exercise. The prima facie standard thus becomes a constitutionally vague terrain.
Furthermore, the investigation window has been increased to a maximum of 180 days under Section 43D(2), from the 90-day ordinary limit under Section 187(3) of the Bharatiya Nagarik Suraksha Sanhita, 2023 before the default bail becomes available. This is double the ordinary period for most serious offences. Upon filing of the chargesheet, Section 43D(5) comes into effect. Since there is no statutory time limit, the accused may remain incarcerated for years. The bail can be refused at each stage, with no institutional mechanism compelling reassessment. InState of Rajasthan v. Balchand it was established that bail is the default position in Indian criminal law, with detention reserved only where the accused poses a risk of flight, evidence tampering, or witness intimidation. This basic rule has been structurally inverted in the case of UAPA
Another flaw is the complete absence of any institutional mechanism to review the burden placed on the accused. Once bail is refused under Section 43D(5), the burden of proof is on the accused to file the successive applications against the same near-irrebuttable standard with no provision requiring courts to periodically reassess whether continued detention remains justified. The parliament is not required to review whether these restrictions remain proportionate because of absence of a sunset clause requiring mandatory periodic renewal of its most coercive provisions. Section 37 of the act establishes Review Committees to hear appeals for denotification of terrorist organisations or individuals under Section 36. However, concerns are raised over the independence of these Review Committees as they are appointed by the executive. Under Section 37(2) and Section 37(3), while the committee is chaired by a sitting or retired High Court Judge, The Chairperson and up to three other members are appointed by the Central Government. This neutralizes institutional safeguards for independent scrutiny by granting absolute powers to the executive in appointments. Thus, the accused faces a dual institutional failure: an unreviewed legislative framework and a structurally biased review standard that fails to provide the independent oversight envisioned by the spirit of Article 22.
Contrast with Comparable Democracies
The contrast with comparable democracies is stark: The United Kingdom’s Terrorism Act 2000 provides for the state to independently re-justify detention before the senior judge at each stage of extension; Canada’s Anti- Terrorism Act 2001 provides a three day cap for preventive detention with mandatory judicial review thereafter; and the United States’s Speedy Trial Act, mandates the prosecution to commence trial within seventy days of indictment or face dismissal of charges. In each system, the institutional burden is on the state to justify continued deprivation of liberty. However, India’s UAPA reverses this entirely, with no sunset clause requiring the Parliament to periodically review whether these restrictions remain proportionate unlike the UK’s Terrorism Act 2006 whose 28 day detention was subject to annual Parliamentary renewal and was ultimately reduced to 14 days after an independent review found it disproportionate, or Canada’s Anti- Terrorism Act whose most extraordinary provisions having never been used in five years were allowed to expire by Parliamentary vote in February 2007 by 159 to 124 against renewal.
Lacunas Used to Incarcerate: From Watali to Gulfisha
In NIA v. Zahoor Ahmad Shah Watali, the Supreme court held that the court must assess the bail applications on “broad probabilities” without scrutinizing the merits. This narrowed the scope for courts to assess the merits of the accusation levelled and provided an upper hand to the prosecution’s material. This left no meaningful opportunity for the accused to contest the chargesheet. This is a deviation from the just, reasonable and fair procedure demanded by Maneka Gandhi v. Union of India under Article 21. This undefined prima facie true standard, assessed from a deliberately light standard of scrutiny of the prosecution’s material, becomes in practice an irrebuttable bar.
In Gulfisha Fatima v. State (NCT of Delhi), bail was denied to certain detained individuals, while five co- accused from the same alleged conspiracy were simultaneously released. The distinguishing label applied was “principal conspirators”. This category is undefined and unrecognized in any judicial precedent. This differential treatment without a statutory basis raises serious concerns under Article 14 of the Indian Constitution. The consequences of these lacunas are visible in the National Crime Records Bureau data which shows the UAPA conviction rates between 1.56% and 6.06% nationally between 2019 and 2023, meaning over 94% of those charged are eventually acquitted. This data was cited by the court in Andrabi while granting bail to the accused. This makes pre-trial detention under UAPA indistinguishable from punishing the innocent.
Contrasting Views by Coordinate Benches
In Union of India v. K.A. Najeeb, the safety valve against indefinite incarceration was introduced. A three- judge bench held that constitutional courts “are not denuded of their jurisdiction” to grant bail under Article 21 when incarceration becomes unreasonably prolonged, and that the rigours of Section 43D(5) “will melt down where there is no likelihood of trial being completed within a reasonable time.” However, this was narrowed down by the two-judge bench inGurwinder Singh v. State of Punjab by ruling that it was a fact specific exception. This is a clear violation of judicial discipline where a smaller bench effectively narrows down a ruling of a larger bench without referring it upward.
Another two-judge bench adopted this reasoning in Gulfisha and held that Article 21 “cannot be a trump card” to override Section 43D(5). This directly inverted Najeeb. Just weeks later, a different two-judge bench comprising Justices B.V. Nagarathna and Ujjal Bhuyan in Andrabi restored Najeeb as binding and held that Gurwinder Singh and Gulfisha Fatima had wrongly narrowed a three-judge bench ruling, and granted bail to an accused held for nearly six years. Hence it is clear that coordinate benches are reaching contradictory conclusions. When the outcome of bail applications depends on the bench and not the law, the rule of law is compromised at its foundation.
Beyond Doctrinal Reconciliation: The Questions the Larger Bench Cannot Avoid
The reference made to the larger bench is much welcome but it will serve its true purpose only when the court answers the following questions- First, what does prima facie true under Section 43D (5) constitute and can it be assessed without providing any adversarial opportunity to the accused? Second, can a court create an undefined category to differentiate between the accused based on gravity of allegations? Third, does the principle laid down in Najeeb call for a time-based presumption in favour of bail? Last, does the absence of a mandatory bail review mechanism violate fair procedure under Article 21?
Suggested Reforms: Balancing detention and liberty
To ensure the balance between detention and liberty, the Supreme Court must initiate the following substantial reforms- First, to ensure the accused has an opportunity to contest the prosecution’s claims, courts should mandate an adversarial review process for the “prima facie” assessment. Second, a time-based presumption of bail must be established in which after a specific duration of undertrial incarceration, the burden of proof shifts back to the state to justify the same. Finally, borrowing from the spirit of Article 22, the Review Committees under Section 37 of the UAPA, must be overhauled to ensure strict independence. These panels must be insulated from executive appointment biases to ensure genuine scrutiny and prevent undertrial incarceration and unilateral individual designations from being weaponised as forms of unconstitutional punishment.
Conclusion
The reference made will be insufficient if the court fails to answer the structural questions raised above and initiate the proposed reforms. With a bail regime marred by ambiguities which squarely affects the fundamental rights of the citizens, the larger bench remains the last forum to provide certainty and clarity.
*The author is a second-year Law student at Hidayatullah National Law University, Raipur. The author may be contacted at advitiya.rai@gmail.com.
This blog reflects the personal views of the author and does not necessarily represent the views of The Policy Chronicle.