Removing the ICC Prosecutor? A Roadmap for What Comes Next

[Ezequiel Jimenez Martinez has a PhD in International Law, is a Senior Fellow at the Center for International Law Research and Policy and an Associated Senior Lecturer at the Pompeu Fabra Law Faculty. He is the author of Governing the International Criminal Court: the History and Practice of the Assembly of States Parties to the Rome Statute (Brill, 2025). All comments and feedback welcome via email, X or Bluesky. 

Sergey Vasiliev is a Professor of International Law at the Open University of the Netherlands and co-editor of Governance of International Courts and Tribunals: Institutions, Practices, and Norms (OUP, 2026).]

On 8 June 2026, after more than eighteen months of proceedings whose novelty defied established precedent at every turn, the Bureau of the Assembly of States Parties decided by a two-thirds majority of members present and voting, to refer the disciplinary case against the Prosecutor to the Assembly of States Parties, to suspend him from duty with immediate effect pending the final decision by the Assembly, and to convene a special session to consider the matter. While the decision remains confidential, the sparsely drafted press release provides a minimum of information and makes explicit reference neither to the “serious misconduct” qualification nor to the Bureau’s recommendation to the Assembly of States Parties.

The Bureau’s decision triggers a legally prescribed chain of events that is itself unprecedented. What follows is not a post-mortem of the Bureau’s conduct and approach to shaping the trajectory of the disciplinary process, though such an exercise is amply warranted and would be crucial for the Assembly to undertake in due course. Instead, it is a forward-looking roadmap of the terrain that States Parties must navigate in the weeks ahead. This piece endeavours to sketch the procedural background and the relevant rules because the ultimate outcome of this protracted disciplinary procedure will hinge upon their interpretation and application. 

A Process That Rewrote Its Own Rules

Before turning to what comes next, intellectual honesty requires acknowledging the complexity and singularity of what has just concluded. The Bureau’s one-and-a-half-year engagement with the disciplinary procedure against Karim Khan was, by any measure, a novelty in judicial governance. It involved the early involvement of the Independent Oversight Mechanism and the subsequent abdication and sidelining of the role it was created to play in such situations; the public disclosure of the identity of the elected official under investigation; the externalisation of the investigation to the United Nations Office of the Internal Oversight Services (an arrangement without statutory basis); amendments to both the Assembly’s Rules of Procedure and the ICC’s Rules of Procedure and Evidence made during the process already underway; the appointment of an ad hoc panel composed of three ‘judicial experts’; the rejection of that panel’s report; the solicitation of new party submissions in apparent conflict with the rules as amended; and, finally, a substantive vote on whether the Prosecutor’s conduct amounted to “serious misconduct” within the meaning of Rule 24 of the Rules of Procedure and Evidence, yet another first in the practice of the Bureau. As evidenced by its decision to refer the matter to the plenary Assembly for final resolution, the Bureau reached an affirmative finding on that last question, and that finding now constitutes the legal predicate for everything that follows.

One must readily concede that an effective governance of the Court sometimes requires a reasonable degree of flexibility and adaptation. Rules cannot anticipate every contingency. But any adaptation must remain anchored in the Statute and in basic legal principles: due process, expeditious proceedings, and deference to independent judicial authority in the determination of legal matters whenever such an authority is engaged. Whether those principles have been duly honoured throughout this process is a question that will require a dedicated institutional review once the present proceedings are concluded.

The Special Session: Convening and Documentation Requirements

The Bureau’s press release indicates it will convene, as soon as possible, a Special Session of the Assembly to place the disciplinary matter before the full membership of 125 States Parties. This would be only the second Special Session in the Assembly’s history. Pursuant to the Assembly’s Rule of Procedure (RoP) 9, a minimum of twenty-one days’ advance notice is required. The agenda must be communicated no later than fourteen days before the session (RoP 14). Given that the Special Session will have a limited agenda, the follow-up on the Bureau’s referral, the logistical lead time does need to exceed a few weeks. Consistent with Article 112(6) of the Rome Statute, the meeting may be held in either New York or The Hague. The former appears the more likely venue, as it would maximise participation by States Parties compared to a special session held in The Hague. The previous special session (7–9 July 2025) was also convened in New York.

What much of the public commentary overlooks is that States Parties are not merely being summoned to cast a vote. The work of the Special Session must be grounded in specific documentation, and the provision of that documentation is not optional . The ASP RoP 81(2) mandates what States Parties shall receive: the investigation report; the underlying evidentiary record; the ad hoc panel’s final report; and the Bureau’s recommendation, being removal from office. Crucially, because the Bureau has elected to depart from the ad hoc panel of judges experts’ findings, it is further required, pursuant to the same Rule and confirmed by ASP RoP 82(2)(c) and ICC Rule of Procedure and Evidence 30(5), to also articulate the legal reasoning for that disagreement. This may be accomplished in a stand-alone Bureau report. 

Such legal reasoning would be of particular significance and should form part of the ‘the relevant information for the purpose of making a decision as to the removal from office’ (see chapeau of ASP RoP 81(2)). The threshold for departing from the considered opinion of three eminent jurists which were the members the ad hoc panel is a high one. Any such decision must be firmly grounded in the legal framework, rather than political accommodation, and it must be reasoned and articulated with sufficient rigor to withstand subsequent scrutiny, including, potentially, before the ILO Administrative Tribunal (ILOAT). Beyond the litigation risk, however, this what the compliance with the rule of law and the proper exercise of disciplinary authority by the ASP and its subsidiary organs require. With view to preserving institutional propriety and the integrity of decision-making, States Parties are entitled to know, before they vote, on what grounds the ASP Bureau chose to set aside the reasoned advice of the judicial experts it itself had selected and appointed.

The Right of Defence before 125 States Parties

A further procedural question awaits resolution before any substantive vote can take place: the right of the Prosecutor to mount a defence. Article 46(4) of the Rome Statute speaks to procedural fairness in removal proceedings. A similar provision is present in Rule of Procedure and Evidence 27(2). It would be proper for the governing body of the Court to ensure that the Prosecutor is afforded the right to be heard before the plenary body that decides his fate. The argument his defence can only take place at the Bureau would curtail the opportunity to make the case to more than just 21 States Parties, who, in the end, need to vote on the matter. 

The suspension of the Prosecutor under Rule of Procedure and Evidence 28, a measure directed at his active exercise of official duties, does not and cannot extend to his basic right of defence. To hold otherwise would conflate the administrative and interim suspension with the deprivation of fundamental procedural rights, a conflation that no robust due process framework permits. The modalities of that defence, whether through an oral address, written submissions, or some hybrid format, remain to be determined by the Bureau in its capacity as the body responsible for agenda-setting and the management of the Assembly session (ASP RoP 15). What the Bureau decides on this point will directly shape both the quality and the legitimacy of the decision ultimately taken by the Assembly.

The Vote Itself: A Two-Step Procedure

The actual voting procedure at the Special Session is itself bifurcated in a manner that has largely eluded public reporting thus far. As prescribed by ASP RoP 63, the Assembly must first determine, by a two-thirds majority of those present and voting (where consensus is not possible), whether the material before it establishes “serious misconduct” within the meaning of the applicable legal framework. This step is confirmed when reading Rule of Procedure and Evidence 29(6) and ASP RoP 82(2)(c). This is Step 1, and it is a substantive legal determination, not a procedural formality. 

Only if that threshold is crossed does the Assembly proceed to Step 2: the vote on removal under  Article 46(2)(b) of the Rome Statute and Rule of Procedure and Evidence 29(4), which requires an absolute majority of all States Parties, namely 63 affirmative votes. It bears emphasis that at Step 1 the Assembly retains the option of finding misconduct of a less serious nature, which under the rules would result in a pecuniary sanction and a reprimand administered by the Bureau, or of finding no misconduct or breach of duty at all, thereby closing the matter. In other words, if the Assembly does find serious misconduct or breach of duty but decides not to remove the Prosecutor, a disciplinary measure consistent with the serious may be applied as per Rule of Procedure and Evidence 29(6), 30(3) and ASP RoP 81(4) and 82(3). If the Assembly decides no misconduct or breach of duty occurred, then the matter is closed as per Rule of Procedure and Evidence 30(6). The two-step architecture is not a technicality; it is the procedural guarantee that removal, a measure of the gravest constitutional consequence, requires a discrete and deliberate collective judgment, not merely a continuation of institutional momentum.

Post-Removal Scenarios: Litigation and Succession

Should the Assembly vote for removal, Rule of Procedure and Evidence 31 stipulates that the decision takes immediate effect. The Prosecutor has stated publicly his intention, in that event, to seek redress before the ILOAT. A significant jurisprudential question arises here: ILOAT case law is calibrated to labour relations within international organisations, and it is far from settled whether a political decision by a plenary treaty body to remove an elected official, an act that is as much constitutional as administrative, falls within the Tribunal’s remedial jurisdiction. The available case law offers limited guidance as to how an award, or any reinstatement order, would interact with the Assembly’s political determination and its institutional and practical implications, which may by that point render the return to office of the elected official found to have been unjustly removed difficult, if not unfeasible. This is territory that the ICC’s governance framework has not previously had to navigate, but it is a question that may become pressing and could linger for some time.

Should removal occur, a fourth election for the position of Prosecutor will be triggered. The history of the three preceding elections provides ample material for institutional reflection. It has become clear by now that States Parties have difficulty adhering to the rules they themselves establish, and this does do not reflect well on the quality of governance they provide to the Court. The 2020 election deviated from the pathway initially agreed by the Assembly, a process that ultimately produced the very officeholder now facing removal. States Parties would be well-served by reflecting critically and honestly on the lessons learned from their previous experiences. There is much work that remains to be done by the ASP with respect to the structural conditions, candidate vetting, campaign conduct, nomination procedures, and the role of the Independent Oversight Mechanism, before the next electoral cycle begins rather than during, let alone after, it.

One further consideration, which should certainly inform, though not overshadow, the next election: the principle of equitable geographical representation, a value embedded in the Rome Statute and consistently invoked by the Assembly. Of the three Prosecutors to date, the office has been held by representatives of Latin American, African, and Western regional groupings respectively. Two regional groups, Asia-Pacific and Eastern Europe, have yet to hold the position. Whether and how this consideration is weighed in the next election will itself be a governance question for States Parties to resolve transparently, according to the Statute’s values rather than the negotiating dynamics of any particular moment.

Conclusion: Policy Choices Under Unprecedented Conditions

What lies ahead for the ICC governance system is genuinely without precedent. Neither the practice of the Assembly nor its accumulated institutional memory provides a ready template for what must now unfold. That absence of precedent is precisely why the quality of the choices made will be so determinative, not only for the immediate question of Prosecutor Khan’s tenure, but also for the safeguarding of prosecutorial independence, as well as for the long-term credibility and integrity of the Rome Statute system. It appears that some of the governance solutions were improvised in a crisis-management mode without a thorough assessment of the legal risks, contingent scenarios, and a clear long-term vision. Such solutions were subsequently refined through piece-meal procedural adaptations devised on the fly rather than through structured and inclusive consultation and carefully considered reform. 

These circumstances warrant serious scrutiny in the institutional review that must eventually follow – for the sake of improving the Court’s governance in the future. The Independent Expert Review process of 2020 provides a sobering precedent for what happens when the Assembly’s reform commitments are not implemented fully with the requisite seriousness and follow-through. States Parties are, in the end, the masters of the Rome Statute system. What the upcoming Special Session requires of them is not the exercise of raw political preference and disciplinary power, but careful engagement with a complex legal record, a rigorous two-step decision-making process, and a keen awareness of the long-term consequences, for the institution, for the principle of accountable and good-faith governance, and for the Rome Statute system as a whole – regardless of how they choose to vote.