The drone threat in 2026

The cheap drone has changed the character of war

The small unmanned aircraft is no longer a niche capability. As of late 2025, Ukraine's ground-forces commander Oleksandr Syrskyi attributed over 60% of all strikes on enemy targets to drones — a share of strikes, not of equipment damage, and lower than the 70–80% casualty-share reported through 2025, with a claimed 96% in March 2026 (video-confirmed; cited publicly by UK Defence Secretary John Healey and Finnish President Alexander Stubb). The economics drive the shift: a first-person-view (FPV) quadcopter costs on the order of $500, yet can disable a vehicle worth several orders of magnitude more. That cost asymmetry, combined with mass production, is why both sides now field unmanned systems by the million rather than the dozen.

Ukraine reports building around 2.2 million drones in 2024 and about 4 million in 2025, with a target near 7 million announced in January 2026 (a stated plan, not an achieved figure; other estimates diverge).

The result is a battlefield structured around attritable mass — systems small, smart, cheap and many, spent in volume and replaced on a cycle of weeks. The threat is also diffusing beyond states: Mexican cartels alone have been tied to at least 221 weaponised-drone incidents since 2021 (per DHS assessments). Understanding the threat means classifying it, and two schemes dominate procurement and doctrine.

How militaries classify the threat

The US Department of Defense sorts unmanned aircraft into five UAS Groups by weight, altitude and speed. A platform jumps to the higher group if any single attribute qualifies (per Congressional Research Service, IF12797):

NATO uses a three-class scheme keyed to weight, with altitude and radius as guidelines (per ATP-3.3.8.1): Class I below 150 kg (micro/mini/small), Class II 150–600 kg (tactical), and Class III above 600 kg (MALE/HALE/strike). Counter-UAS doctrine concentrates on "small UAS" — Groups 1–3, up to 1,320 lb — because that band contains the bulk of the fielded threat. A commonly noted limitation is that Group 3 (55–1,320 lb) is very broad, spanning a hand-launched ISR drone and a long-range one-way-attack weapon in a single bucket.

The threat classes that matter

• FPV quadcopters (Group 1 / NATO Class I). Cheap, precise, and increasingly fibre-optic: a spool of fibre replaces the radio link, so there is no RF emission to jam or detect — fielded fibre ranges run ~15–30 km (prototypes to ~50 km) at a unit cost around $1,200. Electronic warfare, the standard first line of defence, does not work against them: as late as 2024, jamming still neutralised up to ~75% of radio-controlled FPVs — the exact defence a fibre link sidesteps.
• Small ISR drones (Groups 1–2). Low, slow and small, with a poor radar return (radar cross-section ~0.01–0.1 m², bird-to-human) — hard to detect before they have already found the target.
• One-way-attack (OWA) munitions — the Shahed-136 / Geran-2 class (Group 3 / Class II). Public references put the Shahed-136 at roughly 200 kg and about 185 km/h, carrying a 50 kg warhead in the standard variant and around 90 kg in the upgraded 2024–25 variant (per globalsecurity.org, Janes and Army-Technology). These are launched in mass nightly salvos to saturate defences.
• Jet-powered OWA — the Geran-3 / Shahed-238 class. A turbojet variant reported up to ~600 km/h (with observed combat cruising nearer ~300–520 km/h), blurring the line between drone and cruise missile and pushing engagement into higher-end interceptor territory.
• Swarms — mass FPV or coordinated OWA salvos. The hardest problem: they saturate sensors, the decision cycle and the defender's magazine simultaneously.

Why detection is getting harder

The 2026 threat is defined by two moves that defeat conventional defences. First, the spread of fibre-optic control removes the radio link that jammers and RF sensors depend on. The problem isn't theoretical: in 2026 Hezbollah's fibre-optic FPVs repeatedly defeated Israel's Trophy active-protection system — built to stop RPGs and anti-tank missiles.

Second, mass and autonomy mean the attacker no longer needs every drone to survive — saturation is the tactic. Both trends push defenders toward layered, sensor-fused, kinetic answers rather than electronic warfare alone. The scale of the kinetic response is already visible: Ukraine's Mykhailo Fedorov reported in March 2026 that interceptor drones had downed more than 33,000 attacking unmanned aircraft in a single month.

For how militaries answer this threat, see our analysis of the counter-UAS kill chain.