A forensic record exists for what happened in Kosovo in 1999. The graves were uncovered, the chains of command documented, and the judgements entered. What Belgrade’s latest war crimes framing attempts to contest is not the evidence, but the meaning attached to it.
By Kadri Kryeziu, Publisher | 6 May 2026
The bodies came out of the ground at Batajnica in the spring of 2001. Eight hundred of them, in a mass grave on the property of a Serbian Interior Ministry police training centre on the outskirts of Belgrade. They had been trucked from Kosovo in refrigerator lorries during the spring of 1999, sunk in the Danube near Kladovo when one of the trucks gave way, retrieved, reburied. Petrovo Selo yielded more. Lake Perućac more still. The Serbian government's own Working Group documented the operation. The ICTY's Đorđević judgement set out the chain of command. The names of the dead are in the Kosovo Memory Book, the case-by-case forensic record that the Humanitarian Law Center has built across two decades of work.
This is the foundation. Everything else is commentary on it.
I write this because Belgrade's state broadcaster has produced a piece, this past week, on Kosovo's war crimes prosecutions. The piece is built around real data from the Humanitarian Law Center's Pristina office: fifty cases in 2025, one hundred and fifty-four defendants, twenty-four indictments in absentia, ten convictions. The data is correctly cited. The framing around it is not.
The construction has three moves. The first is a complaint about monoethnicity. The defendants in Kosovo's war crimes proceedings are Serbs, and this fact is presented as evidence of prosecutorial bias rather than as a description of who deployed forces in Kosovo during the spring of 1999. The second is the importation of a "Croatian model" comparative frame, supplied by Savo Štrbac of the Veritas documentation centre, which exists for the purpose of producing exactly this comparative frame across post-Yugoslav prosecutions. The third is a bridge to northern Kosovo, supplied by an academic voice, in which the prosecutions are folded into a broader narrative of "state violence against the minority" that includes cyrillic signage cases and university pressures.
Each move works only if the dead are kept out of the room.
So let us bring them back.
Approximately thirteen and a half thousand people were killed in Kosovo between 1998 and 1999. The overwhelming majority were Kosovo Albanian civilians. Forty-four members of one extended family, the Berishas, were killed at Suhareka on 26 March 1999, including children, in a pizzeria into which they had been herded. The bodies were loaded into a truck. Some were later found at Batajnica. At Reçak in January 1999, forty-five Albanian civilians were executed in a ravine, the act that triggered William Walker's emergency report and the diplomatic sequence that ran toward Rambouillet. At Meja in late April 1999, more than three hundred men and boys were separated from a refugee column and shot. At Krusha e Vogël, more than one hundred. At Izbica, ninety-three, executed in a meadow. The list continues. The case files are voluminous. The forensic record is established. The convictions at the ICTY, in Šainović et al. and Đorđević, set out the architecture of command responsibility under which these acts were carried out by Serbian state forces operating under Belgrade's authority.
The Albanians were not killed by Martians.
This is the sentence the RTS piece exists to obscure. The "monoethnicity" complaint is a complaint about the demographic profile of defendants. The relevant demographic profile is the demographic profile of victims, and once that profile is restored to view, the composition of the docket is no longer mysterious. Štrbac's quotable line, that Serbs export plum brandy and Croats export war crimes trials, is the substitution made articulate: defendant-ethnicity offered as the analytical frame so that victim-ethnicity does not have to be discussed. The same logic, applied to Nuremberg, would have produced an objection that the dock was insufficiently diverse.
The four hundred and thirty-four war crimes case files that EULEX transferred to Kosovo's institutions in December 2018 were not generated by ethnic animus in the Pristina prosecutorial system. They were generated by what was done to the people whose remains are still being identified, twenty-six years on, in laboratories that the families of the missing visit because Belgrade has not produced the information that would let them bury their dead. The increased prosecutorial tempo that the RTS piece presents as overreach is the institutional consequence of that backlog finally moving.
The in absentia mechanism, which the piece's Belgrade commentator describes as an "abuse," exists because Belgrade refuses to extradite. The 2019 legal change that enabled in absentia proceedings was a response to the absence of cooperation, not its cause. To object to in absentia trials while defending the non-cooperation that makes them necessary is to object to prosecution itself, dressed in procedural language. This is what the RTS piece is doing. It is not a critique of how Kosovo prosecutes war crimes. It is an objection to the fact that Kosovo prosecutes war crimes at all.
A clarification is worth making here, because the procedural conversation is real. Defence counsel objections to specific in absentia proceedings deserve substantive answers. The families of the missing in Gjakova, in Suhareka, in Drenica have a legitimate stake in how prosecutions are conducted, as do defendants and their representatives. The Humanitarian Law Center's Pristina office is right to publish data that allows scrutiny, and the procedural debate will continue, in Pristina and at the European Court of Human Rights, where it should. None of that is what RTS is doing. The Belgrade objection wears procedural language as cover for a different argument operating underneath.
The Veritas role in supplying that argument is worth naming directly. In practice, Veritas functions less as a documentation body in the general sense than as the institutional vehicle for the comparative frame, in which Serbian prosecutions across the post-Yugoslav space are presented as a coordinated injustice against Serbs, with the demographic composition of defendants offered as evidence and the underlying conduct removed from view. The comparative frame travels because it is portable. RTS imports it for the Kosovo dossier. Republika Srpska affiliates import it for the Bosnian dossier. The architecture is the same. The Štrbac quote is publishable not because it makes an argument but because it documents one being made.
There is a deeper question underneath the news cycle, and it is the question Brussels will eventually have to answer in the Cluster One conversation and in the Chapter 23 file on Serbia's accession track. A state that has not internalised the documentary record of what its forces did in Kosovo during the spring of 1999, that maintains an information ecosystem in which Batajnica is unspoken and the Đorđević judgement is not a settled fact but a contested narrative, that produces broadcast journalism whose function is to keep the dead unattributed, is not a state that has met the values criteria for accession. The "normalisation" framework that the European Union has built around the Belgrade-Pristina dialogue cannot bear the weight of this absence indefinitely. Normalisation without accountability is a structural instability, not a settlement.
What is in dispute, then, is not the procedural debate. The Dispatch will continue to cover that debate as it develops in Pristina and Strasbourg. What is in dispute is the framing Belgrade is offering, through RTS and Veritas and the commentariat that surrounds them, in which the prosecution of war crimes committed against Kosovo Albanians is itself the injustice. That framing is not a perspective. It is memory laundering, with a documentary record, a method, and a network of institutional carriers. Decoding it is part of what this publication exists to do.
The dead are in the ground. The graves are documented. The judgements are entered. The case files are open. And the question of who killed whom, in Kosovo, in the spring of 1999, has an answer.
