Counter-UAS Explained: 5 Ways to Stop a Hostile Drone (2026)

TL;DR: Counter-UAS (counter-unmanned aircraft systems, also written C-UAS or counter-drone) is the set of technologies for detecting, identifying, tracking, and stopping hostile drones. There are roughly five families of defeat methods — RF jamming, GPS spoofing, kinetic interceptors, directed energy (lasers and high-power microwaves), and net capture — and a complete system layers detection (radar, RF, optical, acoustic) on top of one or more defeat methods. The US Army’s Coyote, Anduril’s Roadrunner-M, and Epirus’ Leonidas microwave are the most-visible 2026 examples.
Why read: Counter-UAS is the fastest-growing category in the defense industry and increasingly relevant to civilian critical infrastructure, stadiums, airports, and prisons. AI is the integration layer that makes the whole stack work in real time.
Best for: Anyone tracking defense technology, infrastructure security, or AI applied to physical-world decision-making.
Skip if: You only care about civilian drone photography. Daily AI fundamentals in our free Beginners in AI newsletter.

Hostile drones are no longer a hypothetical threat. Ukraine has shown the world what a $400 quadcopter with a $10 explosive can do to a $5 million tank. The same lesson applies, with different math, to oil refineries, power substations, stadiums, prisons, and presidential motorcades. Stopping drones — quickly, reliably, at scale — is the fastest-growing defense problem of the decade.

The technology that solves it is called counter-UAS (or C-UAS). Here’s how each piece of it actually works, where AI fits in, and which systems are leading the 2026 conversation.

What is a counter-UAS system?

A counter-UAS system is the combination of sensors and effectors that together detect, classify, track, and stop unauthorized or hostile drones. It is not a single device. A real C-UAS deployment is a layered stack:

• Detection layer. Radar, radio-frequency (RF) scanners, optical cameras, and sometimes acoustic sensors that spot incoming drones at the edge of the defended area.
• Classification and tracking layer. Software (often AI-driven) that decides whether the detected object is actually a drone, what kind, and where it’s going.
• Decision layer. An operator console — sometimes with autonomous rules of engagement — that determines whether to engage and with what.
• Defeat layer. The actual mechanism that stops the drone. Five major families, covered below.

The defeat layer gets all the press. The detection and decision layers are usually what actually wins or loses the engagement.

The five ways to stop a hostile drone

Defeat method	How it works	Cost per engagement	Best against	Limitations
RF jamming	Overwhelms drone’s control link with noise on the same frequency	Very low	Commercial drones with operator control link	Useless against autonomous / pre-programmed drones
GPS spoofing	Transmits false GPS signals to redirect the drone	Very low	Drones that rely on GPS waypoints	Useless against vision/inertial-navigated drones; collateral GPS effects
Kinetic interceptor	Another drone, missile, or projectile physically destroys the target	$1,000 – $200,000+	Almost any drone	Cost-per-shot can exceed target cost; magazine depth limits
Directed energy	Laser or high-power microwave damages drone electronics or structure	Pennies per shot (after CapEx)	Swarms, persistent threats, electronic targets	High system cost; weather and power constraints (lasers)
Net capture	Another drone or ground launcher deploys a net	Low	Slow, low-altitude drones near sensitive areas	Short range; not effective against fast or high drones

RF jamming — the cheapest and most-deployed option

The earliest and still most-common counter-drone method. An RF jammer transmits radio noise on the same frequencies the target drone uses for its control link (typically 2.4 GHz or 5.8 GHz) and its GPS receiver. The drone’s receiver can no longer hear its operator, and the onboard flight controller triggers a fail-safe: usually a forced landing, sometimes a return-to-home, occasionally a hover-until-battery-dies.

Notable systems: D-Fend EnforceAir, DroneShield DroneGun, Dedrone HFA, Anduril Pulsar (high-power, vehicle-mounted), Epirus Leonidas (also classified as high-power microwave depending on configuration).

Strengths: cheap to operate, reversible (you can stop jamming and the target drone resumes), proven in Ukraine and at major US airports and stadiums.

Weaknesses: a drone flying a pre-programmed autonomous route doesn’t need an active control link. A drone using visual-inertial navigation (computer vision + inertial measurement) doesn’t need GPS. Both of those modes are now standard on military and on serious commercial drones. RF jamming is the first line of defense, not the only line.

GPS spoofing — redirect the drone instead of jamming it

A more sophisticated variant. Instead of overwhelming the drone’s GPS receiver with noise, the spoofer transmits a higher-power but counterfeit GPS signal. The drone’s navigation thinks it’s somewhere other than where it actually is, and either lands at the spoofer’s preferred location or returns to a fake “home” coordinates.

The most-cited operational example is Iran’s claimed capture of a US RQ-170 Sentinel reconnaissance drone in 2011 by spoofing its GPS to make it land on Iranian soil. Whether that’s exactly what happened has been debated; the technique itself is well-understood.

Limitations are similar to RF jamming: a vision-navigated drone doesn’t care what GPS says. Spoofing also has significant collateral effects — nearby commercial GPS receivers can be affected, which is why this is rarely used in civilian airspace.

Kinetic interceptors — physically destroying the drone

When the threat doesn’t have a control link to jam, or when you need certainty, you need a kinetic engagement. Three categories matter in 2026.

• Interceptor drones. Another drone that flies into the target and destroys it, either by impact, fragmentation, or detonation. Examples: Anduril’s Roadrunner-M ($200k per unit, jet-powered, lands back if not engaged), Anduril’s Anvil (smaller, kinetic), Raytheon’s Coyote Block 2 (kinetic warhead) and Block 3NK (non-kinetic payload, designed for swarms). The Army Coyote has reportedly scored more than 170 combat intercepts as of mid-2026.
• Gun systems and projectile launchers. Traditional small-arms fire is effective against close-range drones; specialized systems like the Northrop Grumman M-LIDS use 30mm chain guns with airburst rounds.
• Missiles. Larger drones (Group 3 and above) are sometimes engaged with traditional surface-to-air missiles — Stinger, Patriot, or in extreme cases F-15-launched air-to-air missiles. The cost asymmetry here is brutal: a $4 million Patriot interceptor against a $20,000 Shahed-136. The reason the Coyote and Roadrunner-M categories exist is to fix that math.

Kinetic systems work against almost any drone but are expensive per engagement, have limited magazines, and produce debris that can itself be hazardous in populated areas. The next two categories address that.

Directed energy — lasers and high-power microwaves

The category every defense ministry is investing in heavily in 2026. Two technologies:

• High-energy lasers (HEL). Concentrated beams of light that heat the drone’s structure or sensors until it falls out of the sky. Cost per engagement: roughly the cost of the electricity to fire the shot. Examples: Lockheed Martin’s HELIOS (deployed on US Navy destroyers), Raytheon’s HELWS, Rheinmetall’s HEL system, BAE Systems’ LaserWeapon, Lockheed’s DEIMOS. Range: typically 1–5 km depending on power. Weakness: atmospheric attenuation (rain, fog, dust), and engagement time on target.
• High-power microwaves (HPM). Pulses of microwave energy that fry the electronics of every drone in the cone of effect. Where lasers engage one drone at a time, HPM can take down a swarm in a single shot. Example: Epirus Leonidas (the leading US system, multiple deployments). Range: shorter than lasers, currently a few hundred meters to ~1 km.

The compelling logic of directed energy: cost-per-engagement collapses, magazine depth becomes practically infinite (limited only by power), and you can engage swarms cost-effectively. The reason it isn’t the dominant solution yet: system cost is high, power requirements are large, weather degrades performance, and the technology is still maturing for some target classes.

Net capture — the lowest-collateral option

For protecting prisons, stadiums, government buildings, and other locations where bringing down a hostile drone with kinetic fire is unacceptable, net capture is the standard answer. Two variants:

• Net guns and shoulder-fired launchers. Ground operators fire a small net at a nearby drone. Effective range a few dozen meters.
• Net-carrying drones. Another drone catches the hostile drone with a deployed net. Some variants entangle the rotors, some carry the target away. SkyWall by OpenWorks is the most-cited commercial system.

Strengths: minimal collateral damage, recoverable evidence (you keep the drone for forensic analysis), generally legal in more situations than other defeat methods. Weaknesses: short range, low success rate against fast or maneuvering drones, useless at altitude.

How AI fits into counter-UAS

The defeat methods above are largely physics and engineering. The AI is in the layers above them.

• Detection fusion. Combining radar, RF, optical, and acoustic data into a single confident track. AI handles the data fusion that no single sensor can do alone — a radar return that’s also showing on RF and on a thermal camera at the same coordinates is almost certainly a drone.
• Classification. Computer vision identifies what kind of drone it is — commercial quadcopter, fixed-wing, loitering munition, decoy. The classification informs the engagement choice. A DJI Phantom at 200 meters is a different engagement decision than a Shahed at 500 meters.
• Threat prioritization. When 30 drones are inbound, which do you engage first? AI models trained on swarm dynamics handle the prioritization decision in milliseconds.
• Decision support. Software like Anduril’s Lattice OS recommends an engagement option — jammer, Roadrunner, Pulsar, manual — based on target classification, distance, and weapon availability. The human operator approves or overrides.
• Autonomous engagement. In specific rule-of-engagement configurations (e.g., a defined geofence around a base, drone classified as hostile by signature), some systems can engage autonomously. DoD Directive 3000.09 governs the rules for autonomous weapons; counter-UAS systems are one of the cleanest cases because the target is itself unmanned.

The strategic significance: AI is what makes layered counter-UAS work in seconds rather than minutes. A human operator looking at three different consoles cannot integrate a 12-drone attack fast enough. A Lattice-style platform can.

Who are the leading counter-UAS companies in 2026?

Company	Flagship counter-UAS product	Category
Anduril Industries	Roadrunner-M, Anvil, Pulsar	Kinetic interceptor + RF/HPM + integration platform
Raytheon (RTX)	Coyote Block 2/3NK, KuRFS radar	Kinetic interceptor + radar
Epirus	Leonidas	High-power microwave
Lockheed Martin	HELIOS, DEIMOS	High-energy laser
Northrop Grumman	M-LIDS, FAAD C2	Integrated kinetic + command-and-control
Dedrone (Axon)	RF detection + jamming	Civilian / venue / corporate
DroneShield	DroneGun, RfPatrol	RF detection + jamming
D-Fend Solutions	EnforceAir	Cyber-takeover (protocol-level, non-kinetic)
OpenWorks Engineering	SkyWall	Net capture

The market is large and growing fast. Industry research firms put global counter-UAS spending in the high single-digit billions in 2026 with double-digit annual growth through the end of the decade. Defense agencies, critical infrastructure operators, and major-event security buyers are the principal markets.

Where is counter-UAS deployed in 2026?

• Forward military bases — The fixed-site layered defense, often Raytheon Coyote + Pulsar + KuRFS combinations, defends bases against drone harassment.
• Mobile / expeditionary units — Smaller deployable systems on vehicles, typically combining RF jamming and short-range kinetic.
• Major-event venues — Super Bowl, Olympics, World Cup, political conventions all routinely deploy counter-UAS layers, usually based on Dedrone-class detection and a defined response protocol.
• Critical infrastructure — Refineries, power substations, data centers, and other high-value sites are increasingly procuring counter-UAS capability. Insurance pressure has accelerated this in 2025–2026.
• Prisons — Drones smuggling contraband (phones, drugs, weapons) into prisons has become a chronic problem. State corrections agencies are deploying RF and net-based systems specifically for this use case.
• Border security — US-Mexico border deployments increasingly include drone-detection capability.
• Airports — The 2018 Gatwick incident put airports on the procurement curve; many major US airports now have at minimum detection capability.

FAQ

What does C-UAS stand for?

Counter-Unmanned Aircraft Systems. Sometimes written counter-UAS, counter-UAV, or counter-drone. The official Department of Defense term is C-UAS.

Why can’t you just shoot down drones with rifles?

Sometimes you can — for close, slow, low drones, small-arms fire works. The problems are accuracy at altitude, magazine depth in a swarm, collateral risk in populated areas, and the basic difficulty of hitting a 12-inch target at 200 meters. Specialized counter-UAS systems exist because rifle fire scales poorly past one or two engagements.

Is RF jamming legal in the US?

Generally no, for civilian use. RF jamming violates FCC rules in almost all civilian contexts. Federal agencies operating in specific authorized roles (DoD, certain Department of Energy facilities, the Secret Service in protective details) have legal authority to use jammers. State and local police generally do not. This is one of the largest legal frictions in expanding counter-UAS deployments.

What is the difference between Pulsar and Coyote?

Pulsar is Anduril’s electronic-warfare counter-UAS family — non-kinetic, primarily RF and high-power microwave. Coyote is Raytheon’s kinetic-interceptor family — a small expendable drone that flies into and destroys the target. They’re often deployed together; one handles cheap commercial drones, the other handles autonomous threats that the first can’t.

How much does a counter-UAS system cost?

Anywhere from $20,000 for a basic RF detection system at a stadium to tens of millions for a fully-layered military base defense. Per-shot costs range from pennies (lasers) to ~$200,000 (Roadrunner-M) to multi-million (traditional surface-to-air missiles). The economic logic of counter-UAS is cost-per-engagement — defeating a $400 drone with a $4 million missile is not sustainable.

Are commercial counter-UAS products available?

Detection products yes; defeat products mostly no. Dedrone, DroneShield, and similar firms sell detection capability to civilian buyers (stadiums, corporate campuses, individual venues). Active defeat (jammers, lasers, kinetic interceptors) is restricted to federal and authorized state buyers in the US.

The bottom line

Counter-UAS in 2026 is not a single product, it’s a layered stack: detection, classification, decision, defeat. The defeat layer has five major families, each with different cost dynamics, range envelopes, and target classes. The AI in counter-UAS is in the data fusion, classification, prioritization, and decision-support layers — not in the missiles themselves.

The single biggest open problem in the field is cost asymmetry. The drone you’re defending against costs $400; the interceptor costs $200,000. Directed energy is the most-credible answer to that math, and the reason every major defense ministry is funding lasers and microwaves aggressively in 2025–2026.

For broader context: AI in Military Drones: The Complete 2026 Overview, Anduril Industries Explained, AI in Drones: The Complete 2026 Guide. Daily AI fundamentals in our free Beginners in AI newsletter.

Sources

• RTX (Raytheon), Critical capability: Counter-drone technologies (February 2026).
• The War Zone, Army Coyote Drone-Hunting Drones Have Scored 170 Combat Kills — press reporting of the 170-intercept figure as cited in this post. The underlying figure originates with US Army operational reporting; treat the specific count as press-reported rather than independently confirmed.
• RTX, Coyote C-UAS product family.
• Unmanned Systems Technology, Directed Energy Weapons — vendor overview.
• IQPC Counter-UAS USA Conference, The U.S. Counter-UAS Market Report 2024–2029 — industry-analyst report. Market-size figures from analyst reports vary substantially across firms; treat them as directional rather than precise.
• U.S. Department of Defense Directive 3000.09, “Autonomy in Weapon Systems” (revised January 25, 2023) — available via the DoD Issuances portal at esd.whs.mil/DD. The governing policy framework for autonomous-defeat rules of engagement, including counter-UAS systems.
• U.S. Department of Defense Joint Counter-Small Unmanned Aircraft Systems Office (JCO) — the DoD organization coordinating counter-UAS strategy. Public communications via defense.gov/News and the JCO’s public-affairs office.
• Anduril Industries, Family of Systems — Roadrunner, Anvil, Pulsar specifications.

You May Also Like

• AI in Military Drones: The Complete 2026 Overview
• Anduril Industries Explained
• AI in Drones: The Complete 2026 Guide
• DJI vs Skydio vs Autel: The Honest 2026 Drone Brand Comparison
• Every AI Model Worth Knowing in 2026