The Trial Judgment in Yekatom and Ngaïssona: A Judgment of Many Firsts Built on Case Management (Part 2)

[Alexander Heinze is an Acting Professor at the University of Bremen and lecturer at the University of Göttingen]

Part 1 of this post examined the Trial Chamber V’s remarkably efficient case management approach in Yekatom and Ngaïssona, highlighting how flexibility became the key to handling an exceptionally complex trial with nearly 20,000 exhibits and 174 witnesses. I explored the Chamber’s innovative approach to modes of liability – particularly its elastic application of Article 25(3)(c) regarding aiding and abetting – and its paradigmatic “submission approach” to evidence that prioritized manageability over strict procedural gatekeeping.

This post turns to the practical application of that flexible methodology, examining how the Chamber’s holistic approach specifically addressed two critical challenges that could have derailed the proceedings: allegations of fabricated evidence and the admissibility of Facebook evidence. These examples illustrate how the Chamber’s deferred admissibility strategy and contextual evidence assessment proved essential in navigating the evidentiary complexities of modern international criminal trials.

Fabricated Evidence in Context

In the course of the trial, the Defence alleged a “deliberate and concerted effort to present false evidence” involving Prosecution witnesses, intermediaries (individuals who facilitate contact between an ICC organ and victims and witnesses, see the full definition in the Intermediary Guidelines, pp. 5-7), alleged former child soldiers, and other individuals, in relation to Count 29, i.e. the conscription, enlistment and use of children under the age of fifteen years to participate actively in hostilities, pursuant to Article 8(2)(e)(vii) ICC Statute (Exclusion Request, paras. 2-3; see also here). The Defence’s case rested on alleged violations of Articles 54(1), 67(1)(e) and (2) ICC Statute, arguing that prosecutorial investigative failures and inadequate supervision of intermediaries created conditions enabling evidence fabrication. Central to their argument was the contention that the prosecution was “wilfully blind to manifest indicia of misconduct and unreliability” and that its “protracted failures” violated fundamental rights while casting “substantial doubt on the reliability of the affected evidence” (Exclusion Request, para. 4). In other words: the Defence argued that the Prosecution knew the evidence was false – not because it factually knew or should have known but it showed ignorance to such an extent that knowledge had to be construed. This was, in the view of the Defence, due to the fact that the Prosecution allegedly failed to investigate obvious signs of fraud and relied on fabricated evidence instead (Exclusion Request, paras. 53-113). TC V did not follow this line of argument, for several reasons: inter alia, it saw no obligation to investigate a general fraud phenomenon, and the alleged conspiracy and investigative failures were only identified retrospectively by the Defence during trial, not at the investigation stage; the Prosecution’s duty to investigate incriminating and exonerating circumstances was limited to the investigation stage, and it was expected to focus on proving charges and had no obligation to ask questions to aid the Defence’s case (Decision on Exclusion Request, paras. 27-8). Even though the latter claims are controversial, they shall not be addressed further – instead, I focus on two other aspects: First, the Chamber was unpersuaded by the “wilful blindness”-construction by the Defence, noting that only the Prosecutor’s (factual) knowledge of the fabrication prevents it from relying on the evidence (Decision on Exclusion Request, para. 29). Since Article 69(7) ICC Statute does not expressly require knowledge of the party that obtained the fabricated evidence through a violation of the ICC Statute, the Chamber refers to the offences against the administration of justice in Article 70 ICC Statute. However, such a restrictive reading of knowledge cannot be derived from the intent provision of Article 30 – which also applies to offences against the administration of justice – nor is it inevitable in connection with Article 69(7) and the question of evidence exclusion in cases of evidence manipulation: in the case of such a fundamental violation of the rights of the accused, it seems appropriate to exclude evidence even in cases of factual indifference. As to the second aspect: In the end, the TC V convincingly did not find a causal link between any violation (of Article 54(1) etc.) and obtaining the allegedly fabricated evidence, so it became irrelevant whether the knowledge component in Article 70 (and 30, respectively) could be read as wilful blindness. Yet, this legal question will surely return in the future. 

Even though the Trial Chamber did not adopt the fabrication narrative, the allegations seem to have informed the specifics of the Chamber’s holistic approach to the evidence assessment. The usual practice of categorising groups of witnesses to facilitate credibility findings as recommended by the Chambers’ Practice Manual (Annex, Guideline for ICC Judgment Structure, para. 18) has been refined: The Chamber began by organising the 98 viva voce witnesses (and the many whose prior-recorded statements were admitted) into three broad analytical clusters—“high-level insiders” (e.g., senior Anti-Balaka or political figures who claimed direct access to command deliberations), “mid-level insiders” (rank-and-file fighters or local coordinators who operated on specific axes of attack), and “crime-base witnesses” (ordinary residents or victims who experienced the violence on the ground) (Judgment, paras. 194 et seq.). Each category, the Chamber stressed, supplies a different vantage-point and therefore a different set of credibility risks: high-level insiders may exaggerate to minimise their own liability or advance political interests; mid-level insiders often “minimise their own involvement” or “distort the facts” to protect comrades; crime-base witnesses can be traumatised or rely on hearsay when describing events beyond their immediate perception. Instead of treating any single inconsistency as proof of fabrication, the Chamber evaluated every witness “holistically” (Judgment, para. 203), cross-checking narratives across categories and against documentary and digital evidence, and explicitly noting when a witness was rejected in toto (P-0487, V45-0001) or only trusted for discrete points. While the word “fabricated” appears nowhere in the operative credibility sections, the Chamber’s method amounts to a step-by-step inoculation against fabrication claims: it (i) refused to elevate speculative links into presumptions of misconduct; (ii) drew nuanced distinctions between mistakes, memory lapses and deliberate deceit; and (iii) clarified that cross-examination can not only enhance the credibility of the testimony but also refute serious allegations of evidence being obtained in violation of the Statute or internationally recognized human rights. 

This is most visible in its treatment of the four “child-soldier” witnesses advanced under Count 29, the allegation of the recruitment of child soldiers (Judgment, paras. 506 et seq.). Having examined forged or overwritten birth certificates, ad-hoc civil-registry extracts and conflicting family testimony, the Judges concluded that the documentary trail supporting the claimed ages of V45-0001, V45-0002, P-2475 and P-2476 was “inauthentic and unreliable”, and that the alleged Anti-Balaka enlistment stories were riddled with chronological impossibilities (e.g., participation in events that occurred after the cited commanders were already dead). The limitations of the Chamber’s holistic evidence assessment become apparent at this point through Judge Chung’s dissent, for whom a contextual approach rendered the inconsistencies “minor” (Judgment, para. 529). The Chamber found the attendant challenges to their reliability “well-founded” and refused to rely on their evidence for any factual finding, let alone to establish the under-15 element of Count 29. Importantly, this scepticism did not remain confined to those four individuals: it fed back into the general credibility calculus applied to Enfants Sans Frontières (ESF) demobilisation lists and the oral evidence of NGO staff whose age assessments had been based on the same compromised paperwork. 

A Milestone in the Assessment of Facebook Evidence

Beyond its nuanced response to fabrication allegations, the Chamber’s treatment of electronic evidence represents another groundbreaking aspect of this judgment. Its handling of Facebook posts, call data records (CDRs), and Ngaïssona’s email correspondence marks a milestone in ICC practice and will certainly serve as a blueprint for future trials. Its treatment is unprecedented at the ICC in both scope and depth. Surely, there were the prior Al-Werfalli case (2017) with the first ICC arrest warrant based largely on social media evidence, and the Bemba et al. case on witness tampering; but the Al-Werfalli case with its seven videos posted to Facebook and other social media platforms never progressed to trial due to Al-Werfalli’s reported passing, and the Chamber in the Bemba et al. case with its four Facebook photographs addressed the issue neither in its Bar Table Admission Decision nor in the final Trial Judgment. TC V in Yekatom and Ngaïssona admitted approximately 943 items without requiring formal Facebook employee testimony, relying instead on the lawfulness of requests for assistance and procedural transmission (Judgment, para. 154). The ramifications of this practice cannot be stressed enough: While the Defence embraced post-truth rhetoric and demanded that the TC “must start from a presumption that all messages are misinformation unless proven otherwise, given that individuals will often present unreliable information as fact”, the Chamber did the opposite and presumed authenticity if material came directly from Facebook via proper channels (Judgment, para. 180). Similarly, for CDRs, records generated routinely for billing and regulatory purposes were presumed authentic when transmitted under lawful procedures and supported by technical witness testimony (Judgment, para. 180). 

Considering the Yahoo email collection associated with Ngaïssona (Judgment, paras. 865 et seq), the Defence argued that processing errors—including inconsistent formatting and incorrect time stamps—undermined the authenticity and reliability of the collection as a whole. Complaints centered on alleged breaches in the chain of custody and the potential alteration of some items prior to disclosure, especially given that only a subset of emails had been authenticated by live witnesses. The TC V, after a granular review of both the digital evidence and corroborating oral testimony, found these criticisms insufficient to taint the Yahoo materials in their entirety: „The Chamber does not find the alleged processing errors or anomalies to be of such a degree that they show a pattern which affects the overall reliability of the collection“, Judgment, para. 869. Ultimately, the Chamber admitted the contested emails into evidence, assessing their weight in conjunction with all available sources and specifically in light of direct testimony contextualizing the exchanges. This approach exemplifies the Chamber’s practice reviewing the evidence as a whole, distinguishing technical irregularities from substantive credibility defects, and reaffirming that the mere presence of processing errors does not, absent proof of actual fabrication, justify setting aside potentially probative documentary material.

Sexual and Gender-Based Violence: The Wait for Justice Continues

The Yekatom and Ngaïssona case is not a showcase for the prosecution of sexual and gender-based violence, as already became apparent during the pre-trial phase (see criticism here and here). The judgment reaffirmed this impression: Some charges related to the rape of a 19-year-old woman during the Anti-Balaka attack on Bossangoa, 5 December 2013. The Chamber affirmed the act of rape occurred but ultimately did not convict Ngaïssona on those counts. That is certainly a disappointing outcome for victims of rape. And yet, it reflects the jurisprudence on the mental element: while the Chamber found that although Ngaïssona had virtual certainty that attacks against Muslim civilians would result in killings and forced displacement, such virtual certainty did not extend to rape (Judgment, para. 4129). Thus, liability under Article 25(3)(c) failed at the knowledge/intent stage for these counts. This is the established threshold (Lubanga, para. 447; Katanga, paras. 776– 7; Bemba et al., para. 29; Ongwen, para. 2695; Al Hassan, para. 1121): the accused must be aware that the consequence will occur in the ordinary course of events, barring unforeseeable intervention; and it excludes dolus eventualis as sufficient for principal or accessorial liability for result crimes at the ICC.

Conclusion

The Judgment in the case against Yekatom and Ngaïssona is a Judgment of firsts: with regard to length; the combination of Judgment and Sentence; the amount of evidence; the assessment of digital evidence, especially Facebook evidence; the way the modes of liability are applied. As a result, its factual and legal complexity has not only been highlighted by the Defence (Request for Extension of Time to File the Notice of Appeal and Appeal Brief, paras. 2, 8) but also by Presiding Judge Schmitt in his announcement of the summary of the Judgment. The Judgment is thus not only an important first step for justice in the Central African Republic (the second step is the case against Seleka commander Said, while charges against another Anti-Balaka leader Mokom were withdrawn in 2023 and the proceedings terminated) but also stands testament to the growing importance of case management decisions in international criminal law. 

Throughout this prolonged trial – spanning over six years from the confirmation of charges to judgment – Trial Chamber V shaped the proceedings through active, efficiency-focused leadership. Whether in the admission and weighing of Facebook posts, call data records, and emails, in distinguishing between memory lapses and deliberate deception, or in the systematic credibility assessment of individual witnesses, it was consistently the procedural architecture that proved decisive. 

What emerges from this approach is an increasingly flexible evaluation style focused on context and plausibility. The Chamber’s “holistic assessment” allows for the integration of complex evidentiary situations without committing to rigid analytical frameworks. This managerial approach extended beyond mere procedural matters into substantive legal determinations, reflecting a broader trend toward applying soft and flexible criteria in international criminal justice. While this openness may be appropriate to the complexity of today’s macro-criminal cases, it complicates targeted engagement with specific weaknesses in the evidence. What begins as nuanced evaluation risks ending in an evidentiary black box where findings on reliability, credibility, authenticity, and probative value blend into an untransparent admissibility analysis that strips the Defence not necessarily of its right but of its ability to dispute evidence effectively.

This insight extends beyond procedural law, as demonstrated by the Chamber’s interpretation and application of Article 25(3)(c): with a normatively unrestricted contribution element and ordinary-course knowledge, the application of the law becomes an overall assessment that is difficult to deconstruct. Not the evidence alone, but its procedural framing decides the result—this too is a lesson from one of the most comprehensive proceedings in ICC history.