Withdrawal as Acquittal

On 23 July 2026, Trial Chamber IV of the International Criminal Court terminated the proceedings against Abdallah Banda Abakaer Nourain and vacated the arrest warrant against him. What makes the decision remarkable is not that a fifteen-year-old case collapsed, but the reasoning by which the judges let it collapse. In the underlying decision of 26 June 2026, the Chamber rejected every argument the Office of the Prosecutor (OTP) advanced for abandoning the case – finding that the evidence still met the statutory threshold for trial, and that trying Banda would work no injustice – and then authorised the withdrawal of the charges anyway, because the OTP simply refused to prosecute. That combination is the tell. A confirmed case that remained, on the judges’ own assessment, triable was extinguished by prosecutorial refusal alone. The termination is in substance a prosecutorial acquittal dressed in the robes of a procedural withdrawal, and it demonstrates that judicial oversight of withdrawal at the ICC does not restrain a determined Prosecutor at all.

A Triable Case

The underlying facts are grave and long-settled. Banda, a commander of a “Justice and Equality Movement” faction, was charged over the 29 September 2007 attack on the Haskanita base of the African Union Mission in Sudan (AMIS) in North Darfur – a situation the United Nations Security Council had referred to the ICC in Resolution 1593 – in which twelve peacekeepers were killed and eight severely wounded. Pre-Trial Chamber I confirmed three war-crimes charges on 7 March 2011: violence to life, intentionally directing attacks against peacekeeping personnel and installations, and pillaging. After Banda, who had initially appeared voluntarily, absconded, the Chamber issued an arrest warrant on 11 September 2014. Announcing the first Haskanita case – against Bahar Idriss Abu Garda – in May 2009, then-Prosecutor Luis Moreno-Ocampo had framed the attack as one on the millions of civilians the peacekeepers had come to protect, and as a serious crime under the Rome Statute that would be prosecuted.

Nor was this a case built on sand. The parties had agreed on a substantial body of facts under Rule 69 of the Rules of Procedure and Evidence, narrowing the dispute to three issues: whether the attack was unlawful, whether Banda knew the facts establishing its unlawfulness, and whether AMIS qualified as a peacekeeping mission under the United Nations Charter. When the OTP sought to withdraw the charges confidentially on 5 October 2023, invoking evidentiary deterioration and exhausted investigative avenues, the Chamber refused on 13 October 2023 – and again, by majority, on 6 March 2024. In November 2025 the OTP renewed its request, submitting that two recently interviewed witnesses could not speak to the planning and execution of the attack or to Banda’s knowledge of its illegal nature, and that it therefore could not prove his mens rea to the requisite standard (paras 9 and 14).

The timing of that renewal deserves more attention than it has received. In October 2025 – a month before the OTP’s final push – the Defence itself had moved to convert the arrest warrant into a summons to appear, submitting that the Government of Sudan was willing to ensure Banda’s appearance for trial; the OTP did not oppose, provided Sudan’s assurances were placed on the record (paras 8 and 17, recording the Prosecution’s response to the conversion request). After twelve years, the fugitive problem was dissolving. It was at precisely this moment, with a trial closer than it had been in a decade, that the OTP chose to abandon the case instead.

The Chamber Wins Every Argument and Loses the Case

The governing framework looks robust on paper. Article 61(9) of the Rome Statute conditions any withdrawal of charges after confirmation on judicial permission, and Regulation 60 of the Regulations of the OTP directs the Office to seek withdrawal where the evidence no longer supports the charges. Article 64(2), under which the Chamber acted, entrusts the trial chamber with ensuring that proceedings are fair and expeditious. The permission requirement reads like a genuine gate: judges, not the Prosecutor, decide whether a confirmed case dies.

What followed shows how the gate operated in practice. The Chamber first disposed of the OTP’s reconsideration request point by point. It held that the new witness accounts were supplementary evidence that did not alter its earlier findings, that the submissions on degradation merely reiterated arguments already rejected, and that the OTP had still not demonstrated how evidence which met the statutory test for trial had deteriorated to the point that a trial was no longer merited (paras 21–23) – recalling that the credibility of witnesses on the disputed issues was for the Chamber to assess at trial, not for the Prosecutor to pre-judge. It then rejected the OTP’s core theory of injustice outright, holding that requiring an accused to stand trial, whatever the outcome, “cannot be regarded, in and of itself, as an injustice” (para. 25).

And then, in the very next paragraphs, the Chamber reversed course. Compelled, as it put it, by its overarching obligations, it noted that the OTP refused to proceed to trial notwithstanding the confirmed charges, the agreed facts, and the opposing views of the victims. The Rome Statute’s hybrid system, the Chamber reasoned, depends on the Prosecution carrying out its responsibilities; a trial with a Prosecutor who will not present the case would contravene the Chamber’s duty to ensure fair and efficient proceedings. The Chamber acknowledged its own statutory powers under Articles 64(6) and 69(3) to call witnesses and order the production of evidence – and declined to use them, reasoning that a chamber driving the case itself, without the Prosecution, would be fundamentally unfair to the accused, and that the victims, though entitled to participate, cannot be assigned the Prosecution’s burden. Deploring the OTP’s stance, it reconsidered its prior decisions and authorised withdrawal to avert what it now cast as an injustice – the very injustice theory it had just rejected on the merits. The Prosecution lost every argument it made and still obtained precisely the relief it wanted. Judicial control that yields at the exact moment it is defied is not control in any meaningful sense.

Kenyatta Is Not the Precedent – It Is the Contrast

The closest analogue confirms rather than softens the point. When Trial Chamber V(B) terminated the case against Uhuru Kenyatta in March 2015, the Prosecution had conceded the evidence was insufficient, and the Chamber had already directed it to choose between withdrawal and trial. There, the judicial and prosecutorial assessments of the evidence coincided; termination followed the merits. In Banda they diverged, and termination followed the Prosecutor’s will over the judges’ contrary view of the merits. Banda thus isolates the pure question – who controls a confirmed case? – and answers it: the Prosecutor does.

The formal ‘without prejudice’ label cannot rescue the characterisation. The OTP has told the Chamber there are no longer substantial grounds to believe Banda responsible; an office that has spent three years arguing that the case cannot proceed is unlikely to revive it. For every practical purpose Banda walks free of the Haskanita charges – without the acquittal on the merits, the reasoned judgment, or the victims’ day in court that a trial would have produced. The victims’ treatment underscores the asymmetry: their Legal Representatives of Victims (LRV) opposed withdrawal throughout, and when the LRV sought leave at the final status conference to challenge the June decision, the Chamber dismissed the request as “untimely and without proper foundation.” The LRV’s mandate now survives only long enough to explain to the victims what was decided over their objection.

Dormancy Is a Choice

The deeper structural failure concerns time. Nothing in the Rome Statute fixes an outer limit on how long confirmed charges may sit dormant against a fugitive while the record erodes. The Haskanita attack has now defeated the ICC twice over: Pre-Trial Chamber I declined to confirm charges against Abu Garda in 2010, the case against co-accused Saleh Mohammed Jerbo Jamus ended with his reported death in 2013, and the confirmed case against Banda has now been abandoned with the judges’ reluctant blessing. Set against the Court’s first Darfur conviction – Trial Chamber I sentenced Ali Muhammad Ali Abd-Al-Rahman to twenty years’ imprisonment in December 2025 – the message to those who attack peacekeepers is that outlasting the evidence works.

Yet the Court has just proven that dormancy is a choice, not a fate. In November 2025, Pre-Trial Chamber III confirmed all thirty-nine charges against Joseph Kony in the Court’s first in absentia confirmation proceedings – advancing a twenty-year-old fugitive case to the very threshold that Banda had already crossed in 2011. The Rome Statute bars trial itself in the accused’s absence, but Banda shows the failure mode is not the absence of tools; it is the absence of any discipline forcing their use. The institutional answer should be a use-it-or-lose-it rule for confirmed charges against long-term fugitives: a fixed period after which the chamber must either set the case down for trial – including via summons where, as here, the accused signals willingness to appear – or dismiss it on a reasoned, merits-based record. Either outcome beats a limbo in which the Prosecutor holds a unilateral veto disguised as a request for permission.

Banda should therefore be read not as procedural housekeeping but as a diagnostic. The confirmation gate swings one way only: judges can stop a weak case from going forward, but they cannot carry a strong one to verdict against a Prosecutor’s refusal. On this evidence, the permission requirement is a courtesy the Prosecutor extends to the bench – not a power the bench holds over the Prosecutor.

The “Bofaxe” series appears as part of a collaboration between the IFHV and Völkerrechtsblog.