Accountability as Instrument: The Death Sentence Against Sheikh Hasina and the Weaponisation of Bangladesh’s International Crimes Tribunal

Arafat Hosen Khan is a Visiting Senior Fellow at LSE Law School, a qualified English Barrister, an Advocate of the Supreme Court of Bangladesh, and author of The Constitution of Bangladesh: People, Politics, and Judicial Intervention (Routledge, 2022).

On 16 November 2025, Bangladesh’s International Crimes Tribunal (ICT) sentenced former Prime Minister Sheikh Hasina to death for crimes against humanity arising from the July–August 2024 uprising. The violence was real, and the demand for accountability is both understandable and necessary. Yet the verdict is better understood as an instance of what scholars of democratic backsliding have called ‘autocratic legalism‘ — the weaponisation of formally legitimate legal institutions to achieve outcomes that law, fairly applied, would prohibit. What makes this case important is that a real institution—one with a statutory foundation dating to 1973—was systematically hollowed out and deployed as an instrument of political retribution, rather than functioning as a neutral court of accountability. The verdict is dangerous because it wears the veneer of institutional legitimacy, and it is internationally significant because that veneer conceals a cumulative pattern of legal manipulation.

I. The Institutional Shell: A Court Without Constitutional Foundation

The structural contamination of the ICT is foundational—it precedes the charges, the evidence, and the verdict itself. Since August 2024, Bangladesh has been governed by an unelected interim administration that amended the International Crimes (Tribunals) Act 1973 (ICTA) four times within eleven months through executive ordinances, without any parliamentary scrutiny. Under Article 93 of the Constitution, ordinances are constitutionally valid only when laid before Parliament and approved within thirty days. With no Parliament in existence, no such approval was ever possible. These amendments—which retroactively expanded the tribunal’s jurisdiction and modified its procedural architecture—are accordingly void. The prohibition on retroactive criminal liability, embodied in the common law maxim nulla poena sine lege and entrenched in constitutional rule of law, is not a procedural nicety. It is the foundational guarantee that state power will be constrained by pre-existing law rather than exercised through law fashioned retrospectively to achieve desired results. Under Article 7(2) of the Constitution, provisions inconsistent with the Constitution are void to the extent of their inconsistency.

The tribunal’s composition compounds the problem. The ICTA requires that ICT judges be drawn from the High Court Division. The reconstituted bench instead included Additional Judges—probationary appointees who had not attained the constitutional rank or security of tenure that the statute mandates. This followed the coerced resignations of legitimate Supreme Court judges, including the Chief Justice, in August 2024, and their replacement with politically aligned successors. Articles 22 and 94(4) of the Constitution guarantee judicial independence and prohibit the arbitrary removal of superior court judges. The combined effect of these manoeuvres is the systematic dismantling of the structural safeguards that give judicial independence its constitutional reality, and the construction of an institutional substitute that wears the form of a tribunal while lacking its legal substance.

II. Process as Performance: Evidence, Procedure, and the Abandonment of Judicial Gatekeeping

A court structurally compromised at its constitutional foundation will not, in the ordinary course, exercise robust judicial gatekeeping over contested evidence. So it proved. The prosecution’s central exhibit comprised leaked audio recordings allegedly of Sheikh Hasina authorising lethal force against protesters—recordings that appear to have been obtained through the National Telecommunications Monitoring Centre (NTMC), a government surveillance body, and that the BBC reported as having been leaked online in March 2025. Article 43 of the Constitution guarantees every citizen the right to the privacy of correspondence and communication, permitting restrictions only when reasonable, imposed by law, and necessary for state security. None of those conditions were established. Article 17 of the ICCPR reinforces this protection at the international level. Yet the tribunal admitted the recordings without any voir dire examination, without requiring the prosecution to establish a legal basis for the interception, and without engaging with the constitutional privacy violation their production appears to have entailed. Here, that function — to stand between the state and the accused and scrutinise what is tendered against them — was abandoned altogether.

The procedural architecture of the trial reinforces this picture. Sheikh Hasina remained in India under an outstanding extradition request; newspaper publication in these circumstances did not satisfy the requirement of actual notice under Article 14(3)(a) of the ICCPR. Her court-appointed lawyer operated without any client communication or instructions—a simulacrum of representation rather than its substance. Lawyers associated with the Awami League faced physical intimidation that precluded effective independent defence work. Cross-examination of prosecution witnesses was prohibited on evidence the tribunal characterised as contradictory, in direct contravention of Article 14(3)(e) of the ICCPR and the right recognised in General Comment 32 of the UN Human Rights Committee to examine adverse witnesses as a minimum fair trial guarantee. Defence counsel was appointed in June 2025; witness testimony began in August 2025; judgment was delivered in November 2025. In a jurisdiction where ordinary criminal trials routinely span a decade, this crimes-against-humanity prosecution was resolved in under five months. Procedural choices that might individually be characterised as regrettable departures from best practice disclose, taken together, a process oriented toward a predetermined outcome rather than genuine adjudication.

III. A Contested Factual Record and the UN’s Evidentiary Assessment

The evidentiary foundations of the prosecution are further complicated by a factual dispute that the tribunal does not appear to have engaged with. The UN OHCHR’s fact-finding report of February 2025 estimated that the July–August 2024 protests may have caused as many as 1,400 deaths. Bangladesh’s own Official Gazette, however, recorded a figure of approximately 834 deaths—roughly half the UN estimate. This discrepancy does not, of itself, invalidate the ICT’s case, though it raises a question that a tribunal concerned with evidentiary rigour ought to have confronted directly:: if the prosecution narrative relies on the UN figure as a proxy for the scale of violence or the degree of state culpability, on what forensic basis was the higher figure adopted rather than the official domestic record? If the casualty record is still genuinely contested, does the prosecution narrative rest on fully tested facts or on provisional estimates absorbed from external reports? These questions are not rhetorical; they speak directly to whether the trial proceeded on a settled factual record, and a defence with adequate resources and time would have been positioned to probe exactly this gap. No such opportunity was ever given.

That concern is deepened by the OHCHR’s own evidentiary conclusion. The February 2025 report expressed confidence that the systematic killings may amount to crimes against humanity—language deliberately calibrated to a probable-cause threshold appropriate for a fact-finding mission’s mandate, not to the criminal standard of proof beyond reasonable doubt. The UN is reported to have concluded that the evidence available was insufficient for criminal conviction at trial. When international fact-finders examining the same evidentiary pool decline to conclude that it supports individual criminal responsibility, and a domestic tribunal proceeds to capital conviction on that material—without engaging with the casualty discrepancy, without conducting a voir dire on illegally obtained recordings, and without permitting meaningful cross-examination—the burden of justification falls squarely on the tribunal. The judgment does not discharge it.

IV. Selective Justice and the Architecture of Impunity

The broader political architecture within which these proceedings were conducted confirms the picture. The Interim Government announced explicitly that those who made the July–August 2024 uprising successful would not face prosecution for acts committed during the relevant period. The Interim Government has granted blanket immunity to regime-aligned participants while pursuing capital charges against the ousted leader.This negates the accountability mandate outright. Article 27 of the Constitution guarantees equality before the law and the equal protection of the laws—a guarantee that selective prosecution directly violates. Amnesty International has long documented the ICT’s structural independence deficit, and Human Rights Watch has raised broader concerns about accountability processes in Bangladesh’s post-uprising transition. The ICT Chief Prosecutor appeared at public political rallies calling for the ban of Sheikh Hasina’s own political party—conduct whose incompatibility with institutional impartiality requires no elaboration. Following the conviction, the OHCHR cautioned publicly that all accountability proceedings, especially on charges of international crimes, must unquestionably meet international standards of due process and fair trial. Selective prosecution corrodes the institution itself: a mechanism that immunises one set of actors while prosecuting only the other extends a political crisis into the courtroom rather than resolving it.

V. The Appeal That Cannot Happen, and the Extradition That May Never Come

The right of appeal in a capital case is not a procedural refinement. Under Article 14(5) of the ICCPR read together with Article 6 (the right to life) and the UN Safeguards Guaranteeing Protection of the Rights of Those Facing the Death Penalty (ECOSOC Resolution 1984/50)—which provide that capital punishment shall not be carried out pending any appeal or recourse procedure—it is a fundamental precondition of the lawful imposition of the death penalty. The ICT Chief Prosecutor has stated that Sheikh Hasina must surrender within thirty days to exercise her appellate right, and that failure to do so causes that right to lapse. The proposition is that to contest her death sentence she must submit to the custody of the state seeking to carry it out. There is no recognised doctrine of international law—whether derived from the International Covenant on Civil and Political Rights, the UN Model Treaty on Extradition, or general principles of criminal procedure—that conditions capital appellate rights on physical surrender in circumstances where surrender is itself exposure to execution. A death sentence from which meaningful appeal is practically impossible is arbitrary by definition, and incompatible with Bangladesh’s obligations under Article 6(1) of the ICCPR not to deprive any person of life arbitrarily.

The extradition dimension adds a further question that the fugitive-forfeiture doctrine entirely fails to engage with. Sheikh Hasina’s departure from Dhaka on 5 August 2024 aboard an army helicopter is widely reported. She has remained in India since, and Bangladesh has formally requested her return under the India–Bangladesh Extradition Treaty of 2013. But India is not automatically bound to surrender her. The treaty expressly permits refusal where the alleged offence is ‘of a political character,’ and commentary on the treaty acknowledges a further discretion to refuse where the request is not made in good faith or is not in the interests of justice. A fugitive-extradition case becomes legally weaker—both under international law and under the political-offence exception specifically—when the underlying prosecution is plausibly connected to regime change, domestic consolidation of power, or the removal of a political rival rather than to neutral criminal justice. The question that India’s authorities must confront, and that the international legal community should also be asking, is this: does the prosecution before a tribunal assembled through constitutionally void amendments, by ineligible judges, on evidence admitted without scrutiny of its legality, in a trial completed in under five months, with appellate rights conditioned on surrender into execution, constitute a request made in good faith and in the interests of justice? The treaty’s own safeguards suggest the answer is far from certain.

The two questions—whether the appeal mechanism functions as a death trap, and whether the extradition request meets the treaty’s good-faith threshold—are not independent of each other. They converge on a single analytical point: the legal architecture surrounding this case has been constructed in a way that makes meaningful judicial review of the conviction practically impossible, whether that review takes the form of a domestic appeal or international extradition scrutiny. That convergence reflects a process designed to be final rather than fair.

Conclusion: Legitimacy as the Precondition of Accountability

The ICT was established in 1973 to address the atrocities of Bangladesh’s Liberation War—a mandate of historic importance and genuine legal legitimacy. As I have argued at the Oxford Human Rights Hub Blog in relation to the interim government’s systematic use of law against political opponents, the pattern that has come to define this administration is one of deploying legal form to defeat legal substance. The death sentence against Sheikh Hasina is the most extreme manifestation of that pattern. The reconstituted ICT was assembled through constitutionally void ordinances; its bench was populated with ineligible judges installed after the coerced removal of independent ones; its central evidence was obtained unlawfully and admitted without scrutiny; its process was conducted without effective notice, competent representation, or meaningful cross-examination; its verdict was delivered at a pace inconsistent with genuine deliberation on a contested factual record; and its appellate mechanism was designed to function as a trap rather than a safeguard. No single one of these defects, in isolation, would be sufficient to render a conviction illegitimate. Institutions of justice, however, earn their legitimacy through the cumulative integrity of the process they produce rather than through isolated best moments.

The deaths of July–August 2024 demand justice. The victims and their families deserve redress that is real and accountability whose authority derives from the integrity of the process that delivers it, not from the political preferences of those who designed that process. Accountability divorced from legitimacy collapses into revenge wearing the clothing of law, and it damages not only the verdict it produces but the institutional capacity for genuine accountability that Bangladesh will need, and may ultimately find it difficult to reconstruct, in the years ahead. The international community — and Bangladesh’s own legal institutions, if they are ever restored to true independence — should recognise this verdict for what it is: the continuation of a political crisis by juridical means, dressed in the language of a criminal trial.