In January 2026, UN member states convened in New York for the first Preparatory Committee session on a binding Convention on the Prevention and Punishment of Crimes Against Humanity. A second session in 2027 will frame the Plenipotentiary Conference expected to conclude the treaty in 2028 and 2029 (UNGA Res. 79/122, 2024). The convention’s substance is being settled in this interval, which makes it worth pausing on a question that has drawn surprisingly little attention: what role, if any, should regional organisations play in the framework the treaty establishes?
The draft articles adopted by the International Law Commission (ILC) in 2019, which anchor the negotiation, answer that question by omission. They reproduce the Rome Statute definition « a widespread or systematic attack directed against any civilian population, pursuant to a State or organisational policy » (ILC, 2019; Murphy, 2020), but situate enforcement almost wholly between States, casting extradition, mutual legal assistance (MLA), and the duty to prosecute or extradite as bilateral obligations. Regional bodies have no place in that scheme. Their absence, this brief argues, is a significant oversight, particularly in light of the African Union’s (AU) leading role in establishing atrocity prevention and response as a regional responsibility. Its institutional development, practical achievements, and persistent shortcomings illustrate both the rationale for including regional organisations in the convention and the conditions under which their inclusion would be effective.
Why the Bilateral Model Falls Short
There is a considered logic behind the ILC’s State-to-State design. As Murphy (2020) observes, it follows earlier instruments such as the Convention Against Torture, holding the focus on State responsibility and leaving the prosecution of individuals to the International Criminal Court; delegations at the Sixth Committee’s most recent session restated the premise that responsibility for investigation and prosecution lies first with national authorities (UNGA Sixth Committee, 2025).
For most offences, the assumption that States will investigate and cooperate is reasonable; for crimes against humanity, it is often flawed. Such crimes are committed pursuant to a State or organisational policy; the very authorities expected to investigate may be implicated in their commission. A system that relies on inter-State cooperation can therefore fail at the precise moment when accountability is most urgently required. When state machinery becomes the direct instrument or sponsor of mass violence, neighboring states are frequently reluctant to initiate bilateral legal proceedings due to delicate diplomatic ties, cross-border economic dependencies, or geopolitical spillover fears. That is the gap a regional layer is placed to close, and no organisation has confronted it more squarely than the AU, whose architecture already supplies what a bilateral treaty cannot: a continental early-warning system meant to register atrocity risk before it hardens, and commissions of inquiry into mass violence, as in South Sudan in 2014, that document crimes no single neighbouring State would take up (ACCORD, 2025). The rest of the AU’s experience is a study in the limits accompanying that reach.
What the AU Built
The AU’s atrocity architecture grew out of a particular failure. Over roughly a hundred days in 1994, some 800,000 people were killed in Rwanda while the Organisation of African Unity didn’t act. The UN Security Council, immobilised for most of the genocide, authorised the French-led Operation Turquoise only on 22 June, by which point the killing was all but finished (UNSC, 1994). A UN mandate made the operation lawful but not neutral. France had been the main external backer of the Rwandan government, and its sudden humanitarian turn was seen in that light. Operation Turquoise saved lives, yet it also let some of those implicated slip away through the zone under its protection (Barnett, 2002; Carlsson et al., 1999). What African leaders took from the episode was that a continent without machinery of its own leaves the vacuum to be filled by others, on terms of their choosing.
Article 4(h) of the AU’s Constitutive Act was the reply, claiming a right of the Union to intervene in a member State confronting “circumstances, namely war crimes, genocide and crimes against humanity”, a provision without parallel in any other regional charter (AU Constitutive Act, 2000). Kioko (2003), the AU’s Legal Adviser during the drafting, casts it as a deliberate response to Rwanda, marking a passage from non-interference to non-indifference. As an ambition, it is unmatched among regional bodies; the authority to invoke it, however, rests with the Assembly of heads of state, so the leaders most likely to fall within its reach are the ones who decide whether it is used. That feature does much to explain the record that followed.
Three Faces of the Same Problem
The distance between the AU’s ambition and its practice shows up in three settings that converge on a single pattern. Consider first the coercive power, tested on only one occasion. As violence intensified in Burundi in December 2015, the AU’s Peace and Security Council authorised a protection force of 5,000, allowed the government ninety-six hours to consent, and, for the first time, recommended that the Assembly invoke Article 4(h) should Burundi refuse (AU PSC, 2015). Refuse it did, and at the January 2016 summit the Assembly withheld endorsement; the episode, one official remarked to the International Crisis Group, had “embarrassed the continent” (ICG, 2016). The instrument built in answer to Rwanda has never been formally used. Its abandonment also exposed a fault line the convention would inherit, for Article 4(h) coexists awkwardly with Article 53 of the UN Charter and its requirement of Security Council authorisation for regional enforcement (Amvane, 2015).
Regional adjudication tells a similar story with the Malabo Protocol of 2014. The AU proposed to give an African court jurisdiction over crimes against humanity, genocide, and war crimes. It was the first regional criminal court ever conceived (Jalloh et al., 2019). A decade later, it remains unratified by all but a few; of the fifteen accessions required for entry into force, Angola supplied the first only in 2024. Its defining provision, Article 46A bis, extends immunity to sitting heads of state and senior officials, and on adoption Human Rights Watch and 140 other organisations denounced the clause as a charter for impunity (De Silva & Amadi, 2024). A court designed for the gravest crimes was thus fitted from the outset to keep its most powerful potential defendants beyond its scope, though its tenth anniversary has lately renewed African-led campaigning for ratification.
The operational record completes the picture from another direction. Where the AU does act, it does so through consent-based deployments such as AUSSOM, its stabilisation mission in Somalia (UNSC Res. 2767, 2024; UNSC Res. 2809, 2025). Recent analysis finds that the host government in Mogadishu has assumed unusual control over the mission’s composition and deployment, which is ordinarily reserved to the AU, bending it toward domestic and regional political ends while leaving the Union, as one official put it: “accountable for an operation it does not direct” (van Emmerik, 2025). Here the AU has not so much shielded a State as been displaced by one, yet the outcome is familiar. In each setting it has built more atrocity architecture than any comparable body, and in each the State reasserts itself wherever that architecture threatens to bind it. Certainly exceptions exist, as when Senegal’s Extraordinary African Chambers convicted former Chadian president Hissène Habré in 2016. However, such exceptions are rare or leaned heavily on outside support.
Comparative Lessons: Inter-American and European Precedents
The systemic tension between regional oversight and state sovereignty is not unique to the African continent. Examining other regional legal frameworks provides vital comparative context for the upcoming draft convention. In the Americas, the Inter-American Commission and Court of Human Rights have pioneered jurisprudence on state accountability for widespread violations, demonstrating how regional judicial bodies can challenge domestic impunity when national systems freeze. The Inter-American Court’s landmark rulings on amnesty laws—such as Barrios Altos v. Peru—invalidated domestic legislative shields designed to protect perpetrators of state-sponsored atrocities, setting an international legal precedent that mass rights abuses cannot be erased by national statutes. Yet, the Inter-American system faces its own structural friction, including chronic underfunding, delayed enforcement, and periodic backlash from member states threatening withdrawal when adverse judgments target high-ranking officials.
Similarly, the European framework under the Council of Europe and the European Court of Human Rights (ECtHR) presents a highly developed architecture for monitoring systemic violations. The ECtHR’s ability to process individual petitions offers a crucial recourse for victims when state machinery fails or actively represses claims. However, the European experience also underlines the intrinsic limits of regional human rights courts: these bodies assess state responsibility and order monetary or structural remedies, but they lack direct criminal jurisdiction to try and imprison individual perpetrators. Furthermore, political resistance to ECtHR rulings in various member states highlights that even the most institutionalized regional mechanisms remain vulnerable to non-compliance when state sovereignty is invoked. These comparative realities demonstrate that while regional systems possess unique normative authority and proximity to local crises, their effectiveness hinges on clear legal integration with broader international standards and robust enforcement guarantees.
What African States Are Asking For
The same tension surfaces in the negotiation when speaking for the African Group at the January 2026 Preparatory Committee, Cameroon welcomed the process and restated a commitment to preventing and punishing crimes against humanity (African Group, 2026, paras. 4, 17). The same statement, though, asked that primacy go to national jurisdictions and bilateral cooperation, that immunities be respected (para. 14), that sovereignty and non-interference be inscribed in the text (para. 12), and that States’ obligations be tied first to national law (para. 15).
A model that elevates national jurisdiction and honours immunities protects exactly those States most likely to commit these crimes, in the circumstances where domestic prosecution is least to be expected; the immunity reflex voiced in 2026 recalls the Malabo “Impunity clause” of 2014 (Malabo Protocol, 2014, art. 46A bis). More recently, Bonita Ayuko’s work (2026) reveals a deep issue within the structural framework of the AU’s court, being that such provision deliberately allows those in power to decide when criminal accountability applies. Though a considerable body of African scholarship makes the case for a stronger regional accountability role in principle, reserving its objection for the immunity clause rather than for regionalisation as such (Jalloh et al., 2019; Kuwali & Viljoen, 2014). The convention must settle the tension in favour of the accountability commitment African States themselves place first, while treating their concerns about sovereignty and capacity as serious rather than incidental.
What the Convention Should Do
The AU’s record points to a practical conclusion: regional bodies have an important role to play in the convention because they can extend its reach where bilateral cooperation falls short, but they will earn that role only if the treaty furnishes the accountability the AU’s own architecture has so often withheld. The most promising starting point is the ground the African Group and this analysis already share: the Group asked for capacity-building and technical support for the States that need them most (African Group, 2026, para. 16), and a regional role built around that request is valuable in itself and the strongest way to make the rest of the case politically durable. Civil societies and regional bar associations’ efforts and cooperation are needed here as well. They play a crucial role in supporting the enactment and operationalisation of implementing legislation for international crimes at the national level. A convention that reserves capacity-building to states only would ignore the longstanding actors who have historically done much of that work.
From there, the least contentious move is one the African Group has all but proposed. A savings clause preserving existing Charter principles and regional frameworks, more specifically the sovereignty and non-interference guarantees the African Group wished to see in the text, would affirm commitments already in force rather than manufacture new ones, on the model of the “without prejudice” language already in the draft articles’ Article 2(3) (Just Security, 2024; ILC, 2019). It asks nothing of a reluctant State that it has not already granted and dispels the fear that a convention with a regional dimension might displace the protections States prize most.
Operationalizing Regional Cooperation: Mutual Legal Assistance and Evidence Repositories
Beyond general capacity building, the convention should explicitly outline structural mechanisms through which regional institutions can facilitate Mutual Legal Assistance (MLA) and cross-border evidence sharing. In many complex atrocities, evidence is dispersed across multiple borders as victims, witnesses, and perpetrators flee, while relevant financial or military records remain locked within regional networks. A purely inter-state MLA regime struggles when diplomatic relations between neighboring states deteriorate or when state officials are complicit. Regional organizations can serve as neutral, centralized clearinghouses for gathering, archiving, and authenticating forensic evidence, digital documentation, and witness testimony. By establishing regional evidence repositories and standardized judicial cooperation protocols, the treaty can empower regional bodies to coordinate cross-border investigations without requiring vulnerable individual states to carry the full political or technical burden alone.
Oversight is the harder and the weightier matter. The draft articles consign compliance largely to States themselves, and Human Rights Watch, among others, has urged that the convention be given a body to monitor it (HRW, 2025; ICJ, 2026). Regional human rights institutions such as the African Commission and the European Court of Human Rights might feed such a review, provided it runs through an independent mechanism and not the body under scrutiny. The AU’s experience supplies the reason: a court that immunised heads of state and a mission captured by its host are evidence enough that regional bodies cannot sit in judgement on themselves. Here the convention should be prepared to press against the African Group’s preferences, since independent oversight is what converts a regional role from a liability into an asset.
The most ambitious proposal is to let regional institutions help bear the cooperation on which the treaty depends. A body such as the AU Peace and Security Council could channel mutual legal assistance among member States that lack the means to manage it alone, again in step with the capacity-building the African Group seeks (ACCORD, 2025). This goes beyond the Group’s insistence on strictly bilateral cooperation and should be presented as such, like a direction to work toward, given that bilateral channels are thinnest precisely where atrocities are most likely. None of these measures creates a fresh obligation: Article 52 of the UN Charter already acknowledges regional arrangements in matters of peace and security, so a convention that recognises them coordinates existing law rather than legislating anew (Radhakrishnan, 2026).
Taken together, these measures draw on the AU’s experience without replicating the institutional ambiguities that have undermined its implementation. The AU is neither a template to be copied nor a constituency whose preferences the convention should ratify, as it is the clearest evidence available of what the regionalisation of atrocity response produces. A reach bilateral cooperation cannot match, shadowed by the readiness of Member States to reassert themselves whenever regional accountability draws near. Those seeking a convention capable of delivering meaningful protection should use the years leading up to the 2028 Conference to advocate for a robust savings clause, an independent monitoring mechanism, and a framework that recognises regional coordination as a complement to individual state accountability. The question of regional involvement will be resolved regardless; what remains at stake is whether the convention helps shape that outcome or merely adapts to it.
In sum, incorporating regional bodies within the future convention is not an argument for diluting state responsibility, but rather for constructing a pragmatic secondary layer of legal defense when state machinery fails. By anchoring regional cooperation within a clear multilateral structure, the international community can ensure that regional institutions act as catalysts for accountability rather than sanctuaries for political expediency.
References
ACCORD. (2025). Evaluation of the African Union Peace and Security Council: Lessons from 20 years of intervention and recommendations for the future. African Centre for the Constructive Resolution of Disputes.
African Group. (2026, January 19). Statement on behalf of the African Group delivered by Cameroon: Preparatory Committee for the United Nations Conference of Plenipotentiaries on Prevention and Punishment of Crimes Against Humanity. United Nations.
African Union. (2000). Constitutive Act of the African Union, art. 4(h).
African Union Peace and Security Council. (2015, December 17). Communiqué of the 565th meeting on the situation in Burundi (PSC/PR/COMM.(DLXV)).
Amvane, G. (2015). Intervention pursuant to Article 4(h) of the Constitutive Act of the African Union without United Nations Security Council authorisation. African Human Rights Law Journal, 15(1), 282–300.
Ayuko, B. (2026, June 23). The “impunity clause” by another name: Article 46A bis of the Malabo Protocol and the crisis of accountability in Africa. A Contrario ICL. https://acontrarioicl.com/2026/06/23/the-impunity-clause-by-another-name-article-46a-bis-of-the-malabo-protocol-and-the-crisis-of-accountability-in-africa/#more-4267
Barnett, M. (2002). Eyewitness to a genocide: The United Nations and Rwanda. Cornell University Press.
Carlsson, I., Han, S., & Kupolati, R. M. (1999). Report of the independent inquiry into the actions of the United Nations during the 1994 genocide in Rwanda (S/1999/1257). United Nations.
De Silva, N., & Amadi, M. (2024). The Malabo Protocol’s 10th anniversary revives advocacy for an African Criminal Court. EJIL: Talk!
Human Rights Watch. (2025). Recommendations for the international convention on prevention and punishment of crimes against humanity.
International Commission of Jurists. (2026). Developing the crimes against humanity convention: Implementation and enforcement of treaty obligations (Briefing Paper 2). ICJ Global Accountability Initiative.
International Crisis Group. (2016). The African Union and the Burundi crisis: Ambition versus reality (Briefing No. 122).
International Law Commission. (2019). Draft articles on prevention and punishment of crimes against humanity, with commentaries (A/74/10). United Nations.
Jalloh, C. C., Clarke, K. M., & Nmehielle, V. O. (Eds.). (2019). The African Court of Justice and Human and Peoples’ Rights in context. Cambridge University Press.
Just Security. (2024, February). Differences ‘getting narrower’ on proposed crimes against humanity treaty.
Kioko, B. (2003). The right of intervention under the African Union’s Constitutive Act: From non-interference to non-intervention. International Review of the Red Cross, 85(852), 807–825.
Kuwali, D., & Viljoen, F. (Eds.). (2014). Africa and the responsibility to protect: Article 4(h) of the African Union Constitutive Act. Routledge.
Murphy, S. D. (2020). The International Law Commission’s first draft convention on crimes against humanity. Case Western Reserve Journal of International Law, 52, 335–404.
Radhakrishnan, A. (2026). The CAH treaty moves forward with first preparatory committee. Just Security.
United Nations General Assembly. (2024). Resolution 79/122: Convention on the prevention and punishment of crimes against humanity (A/RES/79/122).
United Nations General Assembly Sixth Committee. (2025). Summary of the 80th session agenda item on crimes against humanity.
United Nations Security Council. (1994). Resolution 929 (S/RES/929).
United Nations Security Council. (2024). Resolution 2767 (S/RES/2767).
United Nations Security Council. (2025). Resolution 2809 (S/RES/2809).
van Emmerik, C. (2025). The African Union stabilization mission: African mission, Somali rules, Western wallets (DIIS Policy Brief). Danish Institute for International Studies.
