Heligan Report Warns Counter-Drone Defenses Are Starved, But the Government Spending Numbers Tell a More Complicated Story

A consultancy report distributed this week puts a hard number on something drone operators have watched develop for two years: the money flowing into building drones dwarfs the money flowing into stopping them. The report’s author argues that imbalance has left soldiers nearly defenseless against machines that now inflict the majority of battlefield casualties in Ukraine. I have spent the past several months covering one Pentagon counter-drone contract after another, and the spending gap is real in direction. The size of it, and who is doing the measuring, is where this deserves a closer read.

The analysis comes from Heligan Strategic Insights, the research arm of the Heligan Group, and reached a wider audience through the counter-UAS trade portal Unmanned Airspace on June 2. Its central claim is blunt. Autonomous systems now cause up to 80 percent of battlefield losses in the Russia-Ukraine war, and Ukrainian soldiers intercept only an estimated 10 percent of the autonomous attacks aimed at them.

The 80 Percent Figure Measures Two Different Things

Heligan’s headline number describes casualty causation: the share of dead and wounded attributable to drones and other autonomous systems. That is a different measurement from the 80 percent figure Ukraine’s own government publishes, and the two get conflated constantly.

When Defense Minister Mykhailo Fedorov announced at the January Army of Drones event that drones destroy more than 80 percent of enemy targets, he was describing offensive output, the share of Ukraine’s confirmed strikes delivered by drone rather than by artillery or other means. Ukrainian forces logged 819,737 video-confirmed drone hits in 2025 under a bonus-points system that pays units for documented kills, according to Defense News. We covered that announcement in detail. The casualty-causation claim and the offensive-output claim happen to land on the same round number, which makes the conflation easy and the resulting analysis sloppy. Heligan never defines what it counts as a “loss,” but a report arguing for counter-drone urgency should keep offensive output and casualty causation in separate columns, because they answer different questions.

The casualty side does have independent support. One European analysis published in May estimated that 70 to 80 percent of Russian dead and wounded came from drone strikes in 2025. The direction of the data is not in dispute, and it tracks with the production scale we have documented, where Ukrainian FPV output climbed from roughly 20,000 units a month in 2024 to around 200,000 a month in 2025. The precision Heligan attaches to the casualty figure is worth treating with the same caution as any number produced by a firm whose business depends on the conclusion.

Edge Computing Is the Constraint Worth Watching

The strongest part of the report has nothing to do with the casualty statistics. Report author Will Ashford-Brown argues that the real brake on battlefield autonomy was never the artificial intelligence itself. It was running that intelligence fast enough, on the drone, without phoning home to a server.

“The constraint on autonomy has not been AI capability but being able to run this at speed and without relying on a remote server,” Ashford-Brown said in the report. “As edge compute matures, autonomous systems will be able to make real-time decisions and operate more effectively in jammed environments.” That observation matches what combat reporting has shown for months. Electronic warfare and signal jamming saturate the front, and a drone that needs a live radio link is a drone an electronic-warfare operator can kill without firing a shot. Systems that think onboard keep working when the link drops.

This is also where Ukrainian practice complicates the “autonomous weapons” framing. The most effective AI-guided drones in Ukraine do not pick their own targets. A human selects the target, and the drone flies only the final 100 to 1,000 meters on its own, the stretch where jamming and operator error do the most damage. Analyst Kateryna Bondar, in a Center for Strategic and International Studies report, found that autonomous terminal guidance raises strike success rates from 10 to 20 percent up to around 70 to 80 percent, letting a unit finish a target with one or two drones instead of eight or nine. That is a targeting-efficiency story, not a killer-robot story, and the distinction matters for how the threat gets countered.

The Funding Ratio Comes From Sifted, and It Counts Only Venture Money

The report’s most quotable claim is the spending imbalance, and it is worth being precise about where the numbers come from. Heligan attributes them to the venture-capital news outlet Sifted: more than £450 million into UK drone manufacturing since 2019, with drone startups alone accounting for 70 percent of all defense and dual-use funding in 2025, against just £23 million into counter-UAS technology over the same period. That works out to roughly 20 pounds raised for building drones for every pound raised for stopping them.

Two caveats belong on that ratio before anyone repeats it. The figures originate with Sifted rather than Heligan’s own research, and I could not independently confirm them, so treat the 20-to-1 as a claim rather than a checked fact. More important, Sifted tracks venture and private investment, not government procurement, and the government side of the UK ledger looks nothing like £23 million. UK Defence Innovation, the Ministry of Defence body launched in 2025, committed around £30 million to counter-drone technology in its first year alone, inside a wider £142 million drone package. Britain has also signed a £300 million contract to fit DragonFire laser anti-drone systems to Royal Navy Type 45 destroyers by 2027. The private market may favor drones 20-to-1, but the British state is spending on the defensive side at a scale the report’s headline number does not capture.

The offensive side is moving fast too. Ukraine’s largest drone manufacturer, Ukrspecsystems, opened a £200 million production facility in Suffolk in March aimed at building 1,000 drones a month for the front. The gap is real. The 20-to-1 framing oversells it.

The U.S. Numbers Show the Gap Closing Fast

The American picture cuts against the report’s “woefully underinvested” framing, at least on where the spending is headed. The Pentagon’s counter-drone research budget has moved from rounding error to real money in a single cycle. Its joint counter-drone body, Joint Interagency Task Force 401, requested $580.3 million in research, development, test, and evaluation funds for fiscal 2027, up from $6.5 million the year before. That is an 89-fold jump in a single line item.

The same task force already turned that budget into hardware. On May 18 the Pentagon awarded Perennial Autonomy a contract worth up to $500 million for combat-proven drone interceptors, the largest single counter-UAS deal the U.S. military has signed. The threat driving that spending is no longer hypothetical. A small first-person-view drone struck a U.S. Army Black Hawk in Baghdad in late March, and an Iranian one-way attack drone killed six American soldiers at a site in Kuwait on March 1. The threat has migrated so far down the chain that the Army is now building drone-intercept capability into a squad-level grenade launcher. The money is arriving because the body bags arrived first. The “woefully underinvested” line describes the private venture market better than it describes what governments are actually spending.

DroneXL’s Take

Read this report for the edge-computing argument and treat the rest as what it is. Heligan Strategic Insights sells autonomous-systems and counter-UAS consulting. The report reached me through Unmanned Airspace, a portal that runs counter-UAS banner ads and sells a 1,000-product counter-UAS industry directory. None of that makes the spending gap fake. It does mean the people telling you the counter-drone market is starved are the people who profit when it gets fed. I read it the way I read a cybersecurity firm’s annual “threats are worse than ever” report: the trend is usually right, the urgency is always dialed to eleven.

The conflation of the two 80 percent figures is the tell. A firm doing rigorous work keeps offensive output and casualty causation in separate columns, because they answer different questions and come from different sources. Fusing them produces a scarier single sentence and a weaker analysis. I have watched this same number get laundered through three or four outlets since Fedorov’s January announcement, each one shaving off a little more context. By the time it reaches a slide deck it is just “drones cause 80 percent of everything,” which is not what anyone actually measured.

The edge-compute point is the one I would build on. If onboard processing is the real constraint, then the counter-drone problem gets harder precisely as it gets solved, because a jammer is useless against a drone that does not need its link. That reframes the entire detect-and-defeat business that Unmanned Airspace’s directory catalogs. The open question Ashford-Brown raises and does not answer is whether any of the soft-kill electronic-warfare systems that dominate today’s counter-UAS market retain their value once terminal autonomy is standard. JIATF-401’s pivot toward kinetic interceptors like the Perennial platforms suggests at least one major buyer has already concluded the answer is no. Watch the fiscal 2027 defense appropriations process for whether the soft-kill-to-hard-kill ratio in U.S. counter-drone spending confirms that bet.

Source: Unmanned Airspace, citing the Heligan Strategic Insights report and Sifted. Supporting data: Ukraine Ministry of Defence, Center for Strategic and International Studies, UK Ministry of Defence.

DroneXL uses automated tools to support research and source retrieval. All reporting and editorial perspectives are by Haye Kesteloo.