Drone warfare has always been equal parts distressing (it’s impersonal, it’s too easy to leverage, and therefore too easy to abuse) and incredible (ultra-thin fiber as a kite line? Brilliant!) to me. These advances are no different.
That said, I do fret we’re staring down a new age of guerrilla warfare. Drones are cheap, widely available, and increasingly autonomous. Their countermeasures are either impractical for communities (AA Cannons or automated firearms) or costly (jammers, interceptors). The programming can be set-and-forget, meaning operations can be staged months ahead of deployment and make it difficult to find or prevent. The autonomy of target termination specifically raises concerns for the immediate future of violent uprisings, coups, and civil wars.
As an engineer, I am fascinated by it all. As a human, I am horrified that we democratized violence on this scale.
I'm also fascinated by the political implications on state formation, state size, and form of government.
State formation tends to track the relative military effectiveness of large highly-trained standing armies vs. small distributed arms making. The Roman Empire collapsed when they ran out of money to pay their legions. The smaller tribes and kingdoms of the Early Middle Ages unified into the larger kingdoms of the High Middle Ages as the longbow and mounted knight gave the advantage again to large, highly trained standing armies. These collapsed into the city-states of the Rennaissance because the gunpowder musket rendered all the armor of the knights useless. Then the nation-state took over as mechanized arms and airplanes became military weapons, and needed the resources of a large territory to produce them.
It's likely that the drone, being both cheap to produce, easy to use, and extremely lethal to existing weapon systems, will produce a similar political revolution. And it seems tailor-made for smaller political units: drones can lay waste to an invading army, but they suck at power projection because their range is only ~10-20 miles. Might we see a return to city-states as the primary form of political organization? Maybe all the arguments about whether Russia vs. the U.S. vs. China will come out on top are moot, because the very concept of a nation-state will disintegrate, and instead we'll have Beijing vs. Shanghai vs. Shenzhen vs. Moscow vs. Kiev vs. the Bay Area vs. NYC vs. Washington DC? Drones are also ideal for defending shipping lanes, so perhaps we'll see a loose confederation of economically-bound city-states, but each having their own culture and social laws.
IMO the supply chain for drones requires the stability and resource extraction on the scale of a state. SoCs, radios, optics, batteries etc all require high tech manufacturing.
Guerilla use of drones need off the shelf microcontrollers from somewhere, they aren't fabbing them in their backyard.
I would like to see evidence behind this claim on two fronts:
1. all parts are produced in other countries
2. slightly higher prices
There are countries (including Ukraine) that produce on-board flight controllers, but the controllers themselves often rely on components from China. I actually do think it is feasible to create a passable quadrotor using non-Chinese components only but I do not know of a rigorous study or a manufacturer that does this.
I would not provide you with links, sorry. I follow different drone manufactures and over the last years they all had periodic issues sourcing components from China including chips for flight controllers so they found alternatives (i.e. not manufactured in China) for all major components AFAIK .
I've heard that the most difficult to replace is the camera sensor. Drone cameras are produced now in Ukraine as well, however non-Chinese sensors as far as I know are very expensive.
You can source camera sensor from Japan and microcontrollers from France (made at ST Crolles). Your $1 camera becomes $10, $1 micros (s because basic ESC uses 4-5) become $10, $1 Chinese clone IMU becomes $10 TDK IMU, ESC mosfets go from $0.4 to $1.5 (you need 24 of those). Even magnets in the motor are easily doable https://www.kumarmagnet.com/ukraine/neodymium-magnet.html
All the parts are still manufactured in the West/Japan/Korea/Taiwan. Made in China 2025 just means Chinese clones/substitutes are 2-10 times cheaper. In 2015 people laughed and dismissed MIC 2025 as unrealistic, but here we are in 2025 and Chinese products do in fact come with up to 100% domestic electronic components! One stupid example from today, hackaday article https://hackaday.com/2025/06/08/simple-triggering-for-saleae... comment mentions Kingst LA2016. Sigrok wiki https://sigrok.org/wiki/Kingst_LA2016 has pictured of same product at v1.3 and v3.2. v3.2 uses COREBAI clone of Cypress FX2LP and Pango Micro FPGA (https://www.pangomicro.com/en/about/index/) instead of "Intel" Cyclone IV.
"It's likely that the drone, being both cheap to produce"
If you can buy the parts, then yes, but producing a complete drone all on your own is not that easy I think.
Still easier than a stealth air plane or a cruise missile, though. So your predictions might come true, because I also see most state armies being really slow to adopt to this new reality.
> As an engineer, I am fascinated by it all. As a human, I am horrified that we democratized violence on this scale.
How do you even defend against this in a terrorist use case? When a small drone with a grenade or homemade explosive is so accessible? Any Christmas market in central Europe these days is surrounded by car barriers to prevent mass run-overs, but what do you do when soon someone has the idea of dropping some molotov cocktails from drones in public places? Answering my own question I guess you can already throw one manually without a drone. Securing public places is weird, I'm glad it's not my job.
It's going to be an analogue of the situation with firearms today where the assailant has an asymmetrical advantage over civilians but not the countermeasure (police).
Also, the US military has been stockpiling kinetic drone countermeasures for about four years now. The idea is you get a hardened, ~11 pound autonomous drone that slams into the target at roughly 90 mph and physically destroys it before returning home. Add on 1-2 year-old US EW technology that now disables autonomous drones (yep, even autonomous drones), and you can establish a very comprehensive defense. The point I'm making here is that the tech is not only possible, but it exists.
Is it perfect? No. Though defense against firearms and explosives today isn't perfect either. Namely because of response times of the countermeasure. So in that sense, we aren't entering a uniquely dangerous situation.
Edit:
I think what will happen is that the first time a UAS is used on civilians, flying drones around population centers will be banned without permit. That way, if a drone is seen flying without a permit, it gets taken down on sight.
> The fact that we haven't had more drone terror attacks says more about the technological slowness of terrorists than its infeasibility.
Also motivation and incentives. The reason we haven't seen many drone attacks on civilians is that it is far more lucrative to get civilians to buy your product than to kill them, and the companies that actually have the resources to mount a credible drone attack are making a lot more money doing the former.
Terrorism in general has always been far more overhyped than actually a problem - the median number of terrorist deaths per year in the U.S. from 1970-2020 is 4, making your odds of being killed in a terrorist attack significantly lower than being struck by lightning. And the reason is simply that it's deeply irrational. What do you have to gain from killing a random stranger?
non-terror life opportunity
vs
terror life opportunity
Why you tend to have domestic terror in places with bad economies (and high economic equality). And why beliefs often drive it (religious, political, etc.), by adding righteousness to the terror side of the equation.
But requisite minimum skillset is also a consideration, and that's where drones are dangerous.
We're talking (play videogames and some soldering) instead of (chemistry or biology).
The median is more relevant than the mean simply because it's less sensitive to outliers. 9/11 was an outlier; 85% of all U.S. terrorist deaths in the last 50 years happened on that one day.
> How do you even defend against this in a terrorist use case?
You can't, and you don't. That is why it is called terrorism. Safety and freedom are sometimes antagonistic goals. This is an example. Terrorism is defended against by not changing society despite the terrorism. It is violence with a political goal, if the politics do not change, the terrorism fails. Not every soft target can (nor should be) hardened, there will always be soft targets.
Interesting, my take is completely the opposite - you defend against terrorism precisely by changing the society as the material circumstances change, so that the disenfranchised can have their interests represented and achieve a political goal without having to blow stuff up.
(1) Depends on the society. We perhaps agree based on which society we are discussing, and who the terrorists are.
(2) "Defending" soft targets I posit is just an impossible task unless society becomes so locked down that it is unrecognizable. Lots of Sci-Fi has been written on this topic, eg: 'thought crimes', 'judge dredd'.
Sticking to the theme of point (1).. The French revolution was lead by terrorists. They terrorized the aristocracy (and arguably had no other outlet): "cede power or you'll keep getting your heads chopped off." The aristocracy then ceded power. Whether society is forced to change or voluntarily changes is almost an immaterial difference at that point.
OTOH, consider if someone started blowing up libraries in the US. If the response is to shut down libraries, then the population has been terrorized. Bombings in subways were common at a point, if the subways shut down - the population has been terrorized.
In the 1960s, (US) police used dogs against voter registration organizers. To have been terrorized would have been to stop voting, stop voter registration. That terrorism was not successful.
OTOH, lynchings in the US were a prevalent form of terrorism. IE: "this is what happens when you 'don't know your place'". It worked. I recall there is data showing a large change in regional population dynamics for _decades_ following a lynching.
Timothy McVeigh could have done stuff within the libertarian party, but instead chose a path of violence. Timothy McVeigh failed, he blew up a building and that is all that happened.
The 9/11 hijackers chose a path of violence as well. The 9/11 terrorism OTOH was arguably massively successful - a huge wedge developed between cultures & religions, US society radically changed and the US government was successfully baited and began fighting wars that drained its diplomatic, military and financial power.
So, I think for repressed societies, one persons terrorist is another's freedom fighter. In 'open' societies, the terrorist is trying to change the society in some way, through violent means (a shortcut). Changing the society in that way is achieving the aims of the terrorist.
Strictly speaking, the original, “terrorists” for whose “terrorism” in instituting “the Terror” those terms were all coined as specific terms and later genericized were Robespierre and certain other top leaders of the regime established by the Revolution; that groups overlaps with, but is not coextensive with, leaders of the Revolution; the aristocracy already had their legal power stripped before the whole head chopping thing (which was very much not limited to the aristocracy, the vast majority of the ~15k formal executions and the ~10k who died in custody without trial were commoners, and the ratio is even higher for the ~300k killed outside of judicial process during the Terror.)
well, in another case of individual resorted to act thru political violence in New Zealand's Christchurch mass murder... the PM changed the State-wise gun-law but one can also argue that deterrance against 'spreading of terrorism belieaf system' can also be enhanced if the State was to re-enact death sentence for home-grown mass-murderer, no? Soft-targets will always be there and so is mass hysteria, yes?
Thank you for the video! This is so much coming. Plus with thermal cameras too which is even more scary.
On the same topic it reminds me Splinter Cell: Chaos Theory (from 2005!) where an AI called "Masse Kernels" was automatically creating missions sent to different PMCs, managing war assets, supervising the war effort, coms, and the like... Feels also that's coming, at least in some forms for now.
Well you can’t just throw a Molotov because in that case you’re going to get caught. You can place a drone, go home and have it drop the Molotov the next day and no one would know who did it.
Drones aren’t as anonymous. Not sure about other brands, but DJI requires tying it to an email and phone number. It also sends back home a lot of information about your phone, geolocation, network, etc. I’d be surprised if they couldn’t easily tie it to your real life identity.
A few folks on this site might know how to work around these measures. Most of us would probably fail on some subtle way. You’re average terrorist isn’t nearly as tech savvy.
fwiw, I think the car barriers never had the purpose to stop cars, but to show people that at least on some level someone cares about their perceived safety.
At the weekend, I was at a small town which had a half a mile long main road blocked off for a market day. They put up bollards to do so.
Chances are, there were 0 persons planning a car attack on it. So there was an element of "We don't think anything's going to happen, but if it were to happen, we're prepared.". A bit like having a fire extinguisher when there's never been a fire.
But would seeing the bollards also have the effect of discouraging the insane people of the idea of driving the car through the crowds the next time a market day is held?
Oppositely, if they didn't put up any barriers, a psychopath seeing this and the realization that cars can be weapons might give them the idea of "I know what I can do for my act of terrorism..."
I mean…you can’t, at least not right now, not for civvies. Let’s consider some of the current countermeasures:
* Flak/Shrapnel/Birdshot: An excellent last-minute defense if you’re calm enough to line up an accurate shot, but data shows that equipping civilians with these sorts of weapons en masse is a bad idea for safety and well-being. That’s a no-go.
* Nets: Popular for defense, but it’s a matter of time before drones adapt by flying under the nets or changing payload to something to dissolve it. A kamikaze drone could also be enough to destroy an opening for more to swarm. In a civilian context, they’re an excellent deterrent for high-population areas, for now, albeit unsightly.
* Buildings: Safest for now, provided the structure is relatively hardened and the windows are secured. But most civilian structures aren’t guarded against explosions or external attacks, and even those that are require a human to vacate it eventually. Once inside however, there’s more options for stopping an attack - for now - like interior netting, small arms with pellets or buckshot, or even lasers to blind the optical sensors. Impractical for civilian deployment at scale, presently, and highly variable.
* Jammers: Good against piloted drones, but as the article points out, the current crop of dev work is geared towards autonomous slaughterbots instead of human decision-making. Jammers are restricted by most countries and, if left functioning after an attack, could hinder first responders. If left on constantly, would disrupt civilian work. So that’s a no-go.
* LASERS! Probably the best deterrent in the short term for civilians, I would wager. A randomized strobe of a high-powered IR laser could devastate a swarm of drones’ optics, making navigation or target acquisition difficult or impossible. Sticking a piece of protective glass on the sensor would likely nullify it long enough to finish its mission, though.
And that’s what distresses me, ultimately. The future depicted in Slaughterbots or Horizon is rapidly approaching, where autonomous drones can murder with impunity and are affordable enough that any threat actor could get their hands on it. Combined with modern databases of humans - faces, biometrics, profiles, locations, habits, schedules - we’re nearing an era where assassination or murder is a drone away.
That is what horrifies me. And if there’s one thing my time in the defense industry taught me, it’s that nobody is trustworthy with that kind of power. Companies making these absolutely will use them (or condone their use) against dissidents, opposition, regulators, and governments. Pandora’s Box is already open, and I don’t think enough folks appreciate the horrors it will bring.
>Nets: Popular for defense, but it’s a matter of time before drones adapt by flying under the nets or changing payload to something to dissolve it.
Worth noting here that the Ukrainian armed forces have already repeatedly deployed drones with the ability to spray pretty impressive amounts of napalm all over their targets from fairly high altitudes. Fittingly, they've been called "Dragon drones", and I wouldn't want to be under any anti-drone net if one of those arrives.
The "Dragon Drones" also drop thermite [0], a material with massively exothermic high temperature 2000°C+ redox reaction, which can melt through most metals. Thermite is often used to weld rail tracks [1].
Shooting people with a rifle is not scalable and there is a high risk of getting caught. Imagine if school shooters could kill thousands instead of dozens, and at zero risk.
The school shooter type seems to want to be present locally, to go out with a bang, and to do it somewhere they are connected to.
They are not aiming for maximal scalability nor chance to get away.
Of course, other threat actors might, like a Unabomber type or groups designed to destabilize society (perhaps foreign sponsored) by doing repeated actions etc.
I don't want to ban you because you've also posted good things, but if you keep posting ideological and/or political flamebait, and especially religious or nationalistic slurs, we will. Therefore, if you wouldn't mind reviewing https://news.ycombinator.com/newsguidelines.html and taking the intended spirit of the site more to heart, we'd be grateful.
That would be one approach, though I think it is unlikely to be a successful one.
The topic is autonomous drones. Presumably these drones don't stop to go through customs on the way to their targets, and have no reason to respect borders. They're too small for existing radar to track, so enforcement would be difficult.
For that matter, since they can be left in a quiescent state for relatively long periods, there's no requirement that anyone involved be inside the country. They could be shipped internationally through existing shipping services to arbitrary addresses near a target, unpack themselves by cutting through their packaging, and attack a preplanned target.
Literal enemies don't need an invitation. Ukraine was able to smuggle several literal trucks loaded with weapons, deep into literal enemy territory, and put them near airbases where cruise missiles were launched from.
Let's not be naive. Aid programs are a great idea when practical, and I'm certainly not in favor of indiscriminate bombing. But unfortunately in many countries, including Middle East conflict zones, any foreign aid is just stolen by local armed groups and used to fund terrorist activities. How do you think the leaders of Hamas accumulated billions in foreign bank accounts?
Leaders of Hamas has no bank accounts, AFAIK. Did you mean cryptocurrencies? If so, it's unlikely that these funds are personal savings.
Usually, locals doesn't steal from locals in conflict zones, because their throats, or throats of their families, will be cut by other armed locals in no time.
Invaders and mercenaries often rob or kill locals in many, many cases.
Don't be naive. Leaders of Hamas have control of foreign bank accounts. Probably cryptocurrencies as well now. Those funds are the personal proceeds of theft and embezzlement. They steal aid money directly, or steal supplies and sell them for cash. The locals can't do anything to stop them because Hamas has all of the weapons: there are no other armed locals. Hamas violently suppresses the locals who even try to protest against theft of foreign aid.
I don't think any country welcomes extremists like ISIS combatants willingly. So you would need a way to weed them out. It's not just difficult, it's practically impossible.
>How do you even defend against this in a terrorist use case? When a small drone with a grenade or homemade explosive is so accessible?
I can see this from the other perspective. I've read stories for the past 30 years now about police forces and swat teams abusing people, murdering people because someone filled out the address wrong on a warrant, etc. And I wonder, in the coming years, if those sorts of scenarios will be quite as one-sided as they are today. What will that world look like?
>Answering my own question I guess you can already throw one manually without a drone.
Sure, but can you throw one from 2 miles away, can you throw 30 simultaneously? Can you then instantly escape without much of any evidence of who you are being left behind, not even a blurry traffic cam picture? The scale of the mayhem is a quality all of its own.
Speaking just for myself, I am not currently stockpiling explosive-carrying drones in preparation for an encounter with over-zealous police. Yes, such incidents happen. They could even happen to me. Still, I don't worry about them enough to go to that length. (Nor do I see things ending well for me if I did.)
Give it a go and you might be surprised! Stockpiling explosive-carrying drones in preparation for an encounter with over-zealous police can be an educational and absorbing hobby. Many people are sceptical of how things will end for them, but once they start they are drawn in by the friendly and supportive community and doubts like this are quickly forgotten.
Also, the fear of drones is their offensive use. I'm not sure a cop intending to officer-involved-shooting you for looking vaguely like a shoplifter is going to clearly announce his intentions, then wait for you to dig your weaponized drone out of the trunk of your car, boot it up, and get him with it.
Whatever fear Russians have of drones isn't in their offensive use. They're only being used defensively by Ukraine. Similarly, if you're a jackbooted goon, you might well fear their defensive use as I outlined above. But yeh, you have nothing to fear from the defensive use of drones, because you're not attacking anyone... in the coming years, however, the narrative will likely be twisted so that you do come to fear such people as might use them defensively, because the establishment needs you to despise them.
This article did not even cover some of the weird solutions being deployed against drones. For example, Russia has surrounded many of their critical infrastructure sites with huge nets (similar to golf course barrier netting). They have also developed anti-drone drones that drop nets from above, catching and tangling target drones in a bunch of netting that simply snags the blades.
In my opinion, AA guns will become again popular to counter drone swarms. None of the modern weapons are that effective like old AA Guns.India proved this with Pakisthan massive drone attacks recently.
> They have also developed anti-drone drones that drop nets from above, catching and tangling target drones in a bunch of netting that simply snags the blades.
Yeah if we could just all agree that from now on, all warfare is limited to drone-on-drone engagements.
If we could agree on that we'd agree on simply simulating war, no need to actually send drones. But it's unlikely to happen for the same reason hero fights instead of army fights didn't happen like in the movies. Ultimately war is about a difference of opinion strong enough to not care about loss of life, so after a country would lose the simulation they'd just go onto the real war part anyway.
> The programming can be set-and-forget, meaning operations can be staged months ahead of deployment and make it difficult to find or prevent.
This is something I haven't considered before. What's the worst case here? Is it feasible for me to go live on a farm in <country I want to harm>, buy a fleet of DJI drones at flea markets etc, stick something harmful to them, then hide them in the woods.
I can move away, wait a year or two, and then have them fly to the nearest metro area and wreak havoc. This seems to be cheap and relatively straightforward, and hard to detect. What am I missing?
In a pre-drone world you can get explosives, divide them into X equal sized packages, add a timer set to the same point in time to each, then travel around the country hiding them in high-traffic areas.
Yes, that approach is inferior to the drone version. You have to hide them inconspicuously, and a bomb sniffing dog could find them. But you can visit a lot of places in a single European country or US state within one day, and unless the country is already on high alert you can hide something for that time span in public. Yet this doesn't happen. Even regular bombings are rare.
The reasons are manifold: In most places getting explosives isn't actually all that easy (unless you go the homemade route) and is a good method to get attention from authorities. But another factor is that there just doesn't seem to be a large interest in doing that kind of complex attack unless there is already an ongoing civil war. Actual terrorism is fairly rare, and the terrorists tend to be not all that sophisticated.
Are these kinds of drone attacks a scary new possibility? Yes, absolutely! Are they likely to happen? Not really. We might see it as a method to assassinate officials (imagine staging drones at a place where you know the US President will hold a speech in a couple months), but I doubt it will play a major role against the general population
The drone scenario is very different, all you have to do is get close-ish to your target. This is very different than having a stick of c4 sit on your target for months. You have to get them to drone-range, say in the back of a lorry, on a roof somewhere, in a box. Then the programming and or AI can kick in and do the last mile for you, whenever you need. Case in point is the Ukrainian attack on the air fields, they parked a lorry nearby.
Attentive people would find the bombs. Especially if they're on alert, which everyone who knows about terrorism will be at least to some degree.
Back in the day, if you forgot a bag on a British bus the driver would get it and run after you, so that it wouldn't be a bomb issue taking the whole day.
Not much. Autonomous targeting and control are quite new and currently take a fair bit of knowledge and skill to get right, but I expect those barriers to lower dramatically in the coming years. There might be power issues with such a long delay, but I'm not sure. I think the main drawback once this tech gets slightly more widespread is that most people who want to terrorize cities don't want to wait a year or two to do it.
For long running conflicts (Israel vs Iran for example) I expect we'll see some fascinating and horrifying attacks in the near to medium term. Of course anti-drone tech is also evolving quickly and I expect that to continue so the shelf life of any specific attack will probably continue to be rather short.
It doesn't really take much knowledge to set up an autonomous drone mission. It's not DJI-levels of consumer friendliness, but I know multiple people that made and fly their own drones, and it's not something you can't do with a few YouTube tutorials.
They're battery operated, so I think there's a time limit of a few months. Then, you want to be very careful with the infrastructure you leave behind (little pop-open doors/roofs for them to fly out of) to avoid future investigation. You're going to need some practice, your first try will just go to shit perhaps. Opsec while you're setting all this up is still a big deal that amateurs will have trouble getting right. But none of these are particularly insurmountable. With the correct software and careful planning, this will succeed at its goals.
Things you can't help: they will discover the remains of the drones, and also their origin. This evidence will eventually lead back to you (unless you have the aid of a enemy nation-state). Not a big deal if you're dying in a suicide attack, but maybe you don't want the extended vacation in the CIA's worst black ops rendition site.
What you're missing is the will to go through with it. Even state actors would get spooked spending months or years setting this up. One slip-up and you're in prison for life. When your country's existence is at stake, the process is easier.
Battery degradation, a year or two's worth of leaves and debris accumulating on and around the drones, literally all of the elements affecting them, animals, etc.
If I've gone to the hassle of setting up a high tech bombing attack, wouldn't I have also gone to the trouble of putting the drone in a self-opening container?
This no longer sounds cheap or relatively straight forward, it sounds more like an engineering challenge the requires technical expertise and skill to pull off, as well as considerable amounts of money.
You can probably get unsuspecting couriers to deliver last mile. Build a drone dock with cutter to open top of boxes. Mail a weaponized drone to an address within range of target.
Jammers don't work against optical cable or AI vision controlled drones. That's a big problem today in Ukraine for both sides.
As for defense, first of all it's detection and tracking. Copters and long range gas powered drones are very loud and easily detectable. Ukraine uses a net of cell phones. Several devices with microphones can accurately pinpoint all drone like sources in real time. That's cheap to install miles around important targets. Then we need just fast AI interceptors 'on hold', in the air if can afford. The last part is missing today, but we'll get there soon.
As for danger, etc. Small remote controlled firearms were easily available for decades. Drones _are_ trackable. When one takes off in big city Russians know immediately where. By using radio scanners. All DJI drones, and most others, communicate and simply broadcast their coordinates. This is used in Ukraine to find their operators.
What is also distressing is that drones make false flag attacks even easier. Add to that the fact of AI generated media/propaganda means no war will be factually comprehensible to anybody.
I think the idea of a false flag will also be completely destroyed by ai video tech. Once everyone knows everything is fake, how will they know that the false flag was real?
>horrified that we democratized violence on this scale
Violence has always been pretty democratic - you've always been able to punch someone or hit them with a rock and the US seems to have more guns than people.
Pandora's box is now open and multiple groups have access to drones[0].
This is something that I think escapes engineers in this line of work - that something they invented will eventually end up (legally or not) in the hands of people with no scruples.
I'm very interested in the consequences of vigilante applications of this kind of technologies. Imagine a scenario where people start taking out corrupt police officers who have until now been able to terrorize small communities with impunity.
It may motivate actual reform in policing because law enforcement will realize that police officers who kill innocent people with no regard for the law are safer in prison than out on the streets with a paid vacation / desk job punishment.
You need drones to do it in a way that significantly reduces the risk of getting caught.
As it is vigilante action against law enforcement in the west is a sure death sentence and probably life long reprisal against your family once you're dead which is what keeps people in line.
If the development of drone technology significantly reduces the risk of that then you're likely to see many more people respond to violent abuses of authority by law enforcement with vigilante action.
War has (probably) existed since before the dawn of man. Even chimpanzees engage in something like tribal warfare so that behavior probably goes back to a common ancestor species at least 6M years ago.
Also spoofers that could take over a drone - not sure how much encryption is used in most of these off-the-shelf drones, but it would seem like it wouldn't be too difficult to create a Flipper Zero-type device that could spoof the codes used between controller and drone.
That said, I do fret we’re staring down a new age of guerrilla warfare. Drones are cheap, widely available, and increasingly autonomous. Their countermeasures are either impractical for communities (AA Cannons or automated firearms) or costly (jammers, interceptors). The programming can be set-and-forget, meaning operations can be staged months ahead of deployment and make it difficult to find or prevent. The autonomy of target termination specifically raises concerns for the immediate future of violent uprisings, coups, and civil wars.
As an engineer, I am fascinated by it all. As a human, I am horrified that we democratized violence on this scale.