The recent espionage cases point to a coordinated effort not to gather secrets, but to erode Kosovo’s statehood from within.
by Dr. Bahri Gashi | 26 March 2026
Kosovo's Defence Minister Ejup Maqedonci did not speak in abstractions when he addressed the Kosovo Intelligence Agency's recent counter-espionage work. He was precise. "This is the first time since the war," he said, "that such a coordinated and all-encompassing operation has taken place." He saluted AKI director Petrit Ajeti by name. He situated Serbia's intelligence activities within a broader framework of hybrid warfare, one he described as backed by Russia and aimed not at conventional military objectives but at something more corrosive: the delegitimisation of Kosovo's institutions and the questioning of its statehood itself. "Through disinformation and fake news," Maqedonci said, "they aim to portray our institutions as incapable of governing effectively."
The statement was both a public endorsement of an intelligence agency at a moment of operational momentum and a doctrinal description of the threat Kosovo now faces. Both dimensions deserve examination.
What the Record Shows
Maqedonci's characterisation of the operations as unprecedented in scope since 1999 finds support in the case record. Before 2024, publicly reported counter-espionage actions in Kosovo were sporadic and isolated. What has unfolded across roughly eighteen months is different in kind: not the first operations since the war, but the first systematic, multi-case prosecution wave with genuine institutional breadth, spanning intelligence collection, law enforcement, and judicial process simultaneously.
The legal architecture underpinning these cases is Article 124 of Kosovo's Criminal Code, which governs espionage and carries a minimum sentence of five years. Each case follows the same institutional chain: AKI conducts the intelligence collection and surveillance; Kosovo Police executes the arrest; the Special Prosecution of the Republic of Kosovo files the indictment and carries the case through court. That chain, which requires coordination across three distinct institutions with different mandates and cultures, is itself a relatively recent operational achievement. Serbia's Security Information Agency has consistently denied involvement in such networks, framing the arrests as politically motivated. That denial is part of the public record. So is the conviction.
Aleksandar Vlajić, a Kosovo Serb and former Kosovo Police officer, became in 2025 the first person convicted of espionage in Kosovo since the post-war period. He pleaded guilty. The five-year sentence closed one thread. But Vlajić's significance to the broader picture lies less in his own activities than in what his case revealed about the network around him. He did not operate alone. He recruited, managed sources, and served as a transmission layer between local assets and BIA regional and central structures. Among those he allegedly recruited was Jelena Djukanović, a National Programme Officer at the OSCE Mission in Mitrovica.
The Djukanović case illustrates precisely the hybrid dimension Maqedonci was describing. Her alleged value to Serbian intelligence was not access to classified military infrastructure. It was access to institutional process. The indictment charges her not only with forwarding sensitive OSCE documents through Vlajić to BIA but with working to shape the framing of internal OSCE reporting in ways hostile to Kosovo's state institutions. That is a different category of threat: someone positioned inside an internationally credible body, with the ability to influence outputs that policymakers in Brussels and elsewhere take seriously as independent assessment. The Court of Appeals ruled in January 2026 that phone evidence against her is admissible. The trial is ongoing.
The Long Game
If the Djukanović case shows the institutional penetration angle, the case of Hysri Selimi shows something harder to absorb. Selimi, a Kosovo Albanian from Gjilan municipality, is alleged to have collaborated with BIA from November 2009, a period of over fifteen years. The indictment charges him with providing intelligence across an extraordinary range of targets: former KLA members, war crimes narratives, religious radicalisation networks, and Kosovo Albanians who travelled to Syria to fight for ISIS.
That last element requires a moment of stillness. Prosecutors allege that in 2015, on direct instructions from his BIA handlers, Selimi travelled to Syria, joined ISIS, and reported back on Albanian and Kosovar fighters, their numbers, organisation, and areas of operation. He was not a freelancer pursuing his own ideological agenda. According to the prosecution, he was an asset executing a tasking: infiltrate a terrorist organisation, identify your compatriots, report back. He served a Kosovo prison sentence for terrorism-related activity upon return. He then, according to prosecutors, resumed his collaboration with Serbian intelligence.
Intercepted communications released in the course of the investigation identify two BIA handlers by name and show regular contact via Viber, phone calls, and in-person meetings in Serbia. Payments reached up to 700 euros per month. Selimi initially admitted aspects of the collaboration to prosecutors before pleading not guilty in court. Whatever the final judicial outcome, the allegation itself describes something with no precedent in the public record of Serbian intelligence operations against Kosovo: the deliberate deployment of a recruit into a terrorist organisation to surveil Kosovo's own citizens in a war zone.
A Pattern of Targeting
The Selimi and Djukanović cases sit at opposite ends of BIA's apparent recruitment spectrum. What connects them, and what the remaining cases reinforce, is a coherent targeting logic.
A Kosovo Police lieutenant serving as chief of a border checkpoint was arrested in July 2025 on allegations of leaking operational plans and enabling BIA access to official systems. A public commentator arrested in October 2025 is alleged to have transmitted information on Kosovo officials and received materials from a BIA-linked contact for publication. Taken together, these cases describe a programme built around three interlocking objectives: security penetration, to compromise the operational integrity of Kosovo's own institutions; narrative shaping, to feed hostile framings into public and media discourse; and manipulation of international assessment, to influence the outputs of bodies like the OSCE whose conclusions carry independent credibility in Western capitals. The targets are not random. They are the load-bearing structures of a functioning state, selected with precision.
What Maqedonci's Framing Gets Right
Across these cases, a consistent operational picture emerges. Serbian intelligence is not running a programme oriented toward conventional military intelligence collection. It is running a programme of institutional erosion: security penetration at the operational level, narrative shaping in the public information environment, and the manipulation of international assessment through bodies whose independence is assumed. Maqedonci's description of hybrid warfare aimed at questioning Kosovo's statehood is not rhetorical inflation. It is an accurate characterisation of what the indictments allege.
The timing matters. This prosecution wave follows the post-Banjska security environment and coincides with heightened NATO and EU focus on hybrid threats across the Western Balkans. The cases did not emerge in a vacuum; they reflect a regional threat landscape that Western partners are increasingly treating as a shared concern.
The "first since the war" characterisation is therefore as much a statement about Kosovo's own institutional trajectory as it is about Serbian intelligence activity. AKI was established only in 2009. The prosecution infrastructure capable of handling espionage cases of this complexity required years beyond that to mature. What the 2024 to 2026 wave represents is the first moment when collection capability, prosecution capacity, and judicial process were simultaneously operational and coordinated at a level adequate to the threat.
That is a meaningful threshold. As with all counterintelligence work, much of the underlying evidence remains non-public and cannot be independently verified beyond court proceedings. But the proceedings themselves are a matter of public record. The cases are the evidence. The persistence of the activity is the signal that the contest is ongoing.
Dr. Bahri Gashi is a professor of security studies at UBT in Prishtina. His work focuses on intelligence oversight, national security, and cybersecurity in the Western Balkans.
