A UK defense think tank has catalogued 144 drone sightings across a dozen NATO members and Ireland between August 2024 and February 2026 and traced most of them to commercial vessels in Russia's shadow fleet, a sanctions-evading armada of aging tankers that, in the IISS assessment, may now be doubling as low-cost launch platforms for surveillance and harassment flights over Western Europe.
The International Institute for Strategic Studies (IISS) published the synthesis on July 3, 2026, arguing that the geography and timing of the sightings (half over military bases including sites hosting US nuclear weapons, about a quarter over critical infrastructure such as ports and energy hubs, and roughly a fifth over civilian airports) are "consistent with" a coordinated Kremlin probe rather than a coincidence of hobbyists, spooks, or Ukraine-war spillover (Ars Technica, The Aviationist).
Russia's shadow fleet is the loose network of commercial vessels, many aged, poorly insured, and operating under opaque ownership, used to move Russian crude past the G7 price cap and EU embargo. AIS tracking data places these ships, plus other Russian-linked commercial traffic, loitering or transiting in the same waters and approaches as roughly 70% of the drone incidents IISS reviewed. A vessel's cargo deck functions as a launch pad; its cargo is sanctioned oil. Firing up a small commercial drone from a ship deck requires no special infrastructure, leaves no obvious launch signature, and the launch vessel itself looks like ordinary maritime traffic on AIS (Unmanned Airspace, TVP World).
In February 2026, Sweden jammed a drone that had just taken off in its waters from the Zhigulevsk, a Russian Navy signals-intelligence vessel. Sweden's armed forces published the interception. The Zhigulevsk is one ship in 144 incidents, but the only one with a confirmed launch chain (The Guardian, ABC News).
Most of the 144 flights happened at night or in the pre-dawn window. Witnesses and air-traffic controllers described the drones as "professional" or "military-style" rather than consumer quadcopters. The targets, including air bases in Belgium, the Netherlands, Germany, the UK, Poland, and the Baltics (several of which host US B61 gravity bombs under NATO nuclear sharing), liquefied-natural-gas terminals in the North Sea, and refineries in the Baltic, sit exactly where an adversary would want to map detection, identification, and escalation timelines (The Insider).
NATO's air-defense architecture was built for manned fast jets and ballistic missiles, not for slow, low, off-the-shelf quadcopters launching from a moving ship 12 nautical miles out. Identifying a $400 drone costs the same in flight hours, radar time, and operator attention as scrambling a Typhoon. Shooting one down over a populated base risks collateral damage worse than the drone itself. Each incident consumes defender capacity. It does not give the defender a casus belli. Whether any individual flight was Russian, accidental, or a decoy matters less than that the aggregate pattern produces the same intelligence on the defender's response posture that a more aggressive probe would.
IISS does not claim all 144 sightings were Russian drones, or even that all 144 were real drones. The think tank uses "likely," "suspected," and "consistent with" throughout. AIS correlation places shadow-fleet hulls near the incidents; it does not prove they launched the airframes. Moscow has not publicly acknowledged the campaign; Russian embassies in Europe have called the drone reports hysteria (Ars Technica).
European interior ministries have spent 18 months chasing individual sightings. Counter-UAS units exist in most NATO air forces but operate under rules of engagement that assume a sovereign airspace, not a launch from international waters. The next tracked trigger is whether NATO formalizes a maritime-launch counter-UAS doctrine at the Washington summit later this year, and whether the shadow fleet, which has already adapted to oil sanctions, adapts again.