My question was inspired by that one on robot police

Imagine a world where police is mostly operated by robots, at least for the simple operations that does not need high-level decisions. Of course, like in any other worlds, there is a technology race between the warrants of public order and the criminals.

How will criminals adapt and how will they use robots for their criminal activities?

Will they use it mostly as weaponry? Use hacking robots? As a substitute ("go rob the bank")? Have they smarter ways to use robots, for exemple to commit new types of crimes that the robotic technology would allow?

For the sake of this question, robots need not to be humanoid, so drones apply. But are autonomous: they don't need human supervision, at least no more supervision than human police in our world. They can analyse a simple situation, and follow simple orders. For example, the effective operations of anti-riot force and SWAT-like units will be operated by robots, but the strategy will be decided by humans.

  • 4
    $\begingroup$ They will adapt just fine, thank you. By and large, criminals do their criminal work in such a way as to avoid meeting the police. That the police are men, robots or martians is irrelevant. $\endgroup$
    – AlexP
    Dec 24, 2016 at 13:26
  • 3
    $\begingroup$ What technological level are your robots at? A bank robber will hardly be scared by a roomba. $\endgroup$ Dec 24, 2016 at 13:26
  • 1
    $\begingroup$ @JanDvorak: I was thinking a similar level than in the motivating question. I have difficulties to imagine a "police force mostly operated by Roombas" ;). I'll edit to try to precise the question though. $\endgroup$
    – Taladris
    Dec 24, 2016 at 13:33
  • 1
    $\begingroup$ Robo-Cop: "Mdm please drop the EMP now you have 30s to comply, 29..." ,8yo girl: "mummy..."[start crying], Crook: "Boss your plan is working are u seeing this live streaming on YouTube but so sorry kid..." $\endgroup$
    – user6760
    Dec 24, 2016 at 14:03
  • 2
    $\begingroup$ How can you (a member of the public) tell if a robot's a police robot or a fake? I'd create a robot that appeared as much like an official robot as possible, then have that subtly abuse /misuse its power. $\endgroup$
    – JohnLBevan
    Dec 26, 2016 at 10:09

5 Answers 5


I think the most likely, or at least most efficient way for criminals to adapt, and have no doubt they will, would be for them to start with hacking.

It doesn't necessarily even need to be overt control or an obvious hack that alters the day to day function of the police bots.

Criminals would be best served by hacking them and gaining access to the police network. With this access criminals could stay one step ahead of the police and ensure they have a heads up when police actions are imminent.

This would give them time to prepare or better yet relocate and avoid the police altogether.

It would also allow them to gather data on routes and coverage to better plan their nefarious deeds, and to avoid detection and capture.

Now, in the event that things get out of hand or a police operation takes place with regular old humans criminals could wait for an opportune moment and take control of the bots to throw the police into chaos and start infighting so as to cover an escape or turn the tide of a fight.

Police bots seem like they may be a bad idea...

  • $\begingroup$ Would hacked policebots correspond to (human) cops taking bribes? Plus ça change... $\endgroup$ Dec 27, 2016 at 1:15

Crime in a Robot-Friendly Setting

The ability to field a police force with robots implies advances in AI, telecommunications, and related fields. That means the digital economy is the economy. Only hopeless drug addicts go mugging people in the physical world, any successful criminal will do it in the digital world.

Advances in surveillance tech might mean that any criminal has to take much greater care to protect their tracks. Physical involvement with the victim is right out. The digital trace has to be laundered through "darknet" services, servers in Nigeria and Russia and some weird little possession of the British Crown, and so on.

Specific Reactions to Robot Cops

  • How heavy are those thingies? Set your headquarters in run-down housing where the floor can be prepared to break under the weight of a robot. The crashing noise gives enough warning to take the prepared escape route across the roofs.
  • Which senses do they use to track people? Can they be fooled? For instance, can there be face masks that are designed against a (known) face recognition algorithm? Early systems could be thrown off by glasses.
  • Is it possible to track the deployment of police robots? Do they have distinctive commo signatures that could be pattern-matched?

You zap all the police robots and drones with an EMP drone.

All the delicate circuits within the machines will short out after a single burst of an EMP. Some might think ElectroMagnetic Pulse can only be created by nuclear weapons, but militaries have created non-nuclear EMP's with cruise missiles and drones.

Imagine drones that fly over a city shorting out all the non-hardened computer equipment everywhere. Every machine in the city shuts down. All the criminals can get away with anything for a few hours until power is restored and police deploy any drones and robots that were in storage. (Think of The Purge story.) The criminals will know in advance when the Pulse occurs, but everyone else will go about their lives unsuspecting until all hell breaks loose. The criminal gangs will store their cars and their computers inside concrete basements until after the Pulse is over.

Everyone else will see their cars stop suddenly, their smart phones die, their laptops die, and their bank security systems fail. Criminals could just drive where they want, grab what (or whom) they want, and kill whom they want. Nobody's smartphone camera could record their crimes. The criminals would know not to use cell phones after the Pulse. They would

When the police finally deploy their backup robots from storage, the criminals use their many drones to Pulse again and again until all police drones and bots are zapped. Then criminal gangs will reign supreme for days until the governor or president sends in the National Guard to restore order. By which time, the criminals have left town.

To quote Wikipedia on the topic of non-nuclear EMPs:

The range of non-nuclear electromagnetic pulse bombs is much less than nuclear EMP. Nearly all such devices used as weapons require chemical explosives as their initial energy source, producing only one millionth the energy of nuclear explosives of similar weight. The electromagnetic pulse must come from within the weapon, while nuclear weapons generate EMP as a secondary effect. These facts limit the range of NNEMP weapons, but allow finer target discrimination. The effect of small e-bombs has proven to be sufficient for certain terrorist or military operations. Examples of such operations include the destruction of electronic control systems critical to the operation of many ground vehicles and aircraft.

It would take a fairly sophisticated criminal organization to create and deploy EMP drones, but it is possible. Somebody with knowledge of physics, chemistry, and electrical engineering could do it. You would also need to channel a huge amount of power onto the drones. It's also likely each drone would be a one-and-done device. If you use a chemical reaction to generate the energy for each Pulse, that reaction would destroy the drone. Which is why you think of these drones as bombs, and not something you can fire again and again like a gun.

The criminals could only do this a few times until police harden their drones and bots against EMPs, and corporations store all their major computer hardware deep inside concrete bunkers in the bottoms of buildings. And the people would know the next time a Pulse occurs, the governor will send in the National Guard immediately to shoot on sight.


I hate to say this when you've put so much thought into the question, but basically by ignoring the matter.

Unless you've equipped these robot officers with marijuana detection sensors, they're going to be less effective on the street and unable to run a complex investigation.

The average idiot on the street being pulled over for running a red light or being over the speed limit might be more likely to get in trouble, but robots probably won't suffer from prejudice and, unless programmed to, will probably not profile people. This means that the rate of casual malicious enforcement is likely to drop. When told to random sample, they will pseudo-random sample, they won't randomly sample all the Arabs or Africans, they won't randomly sample everyone with dreads.

But the real professional criminals? That needs high level decision making, cross force communication, international co-operation. Investigative skills, human contact skills, and usually a great big load of luck. Robot law enforcement aren't going to be interacting with real criminals.


"It takes a thief to catch a thief."

I don't know who coined it, but the saying seems generally accepted as true, and highlights a commonly perceived obstacle in law enforcement, that a law abiding officer -- which, I must assume, is one reason to have a robot police force, so ideally human sin and error are minimized -- however skilled, simply does have a criminal mindset. Professional criminals perceive the entire world differently, so adaption will likely be in unexpected, criminally brilliant ways.

However, there are also criminals who inadvertently outwit the law, and by unconscious heuristics subvert the system, and not by being very intelligent, but intuitively or by plain luck.

For example, if robot/people cops only use the newest and best technology, and the fortunate criminal keeps using his or her legacy system for communication, either by chance or some belief that it's 'lucky,' such a criminal would be very difficult to catch, especially in an advanced society. So long as no major blunders were made at least. Another way of saying this is, what if it never occurs to cops in a future society to read snail mail?

I think two things from Murphy's Law of Combat are parallel to the criminal mindset which cause untold trouble to law enforcement, which make criminals difficult to handle, regardless of technology:

  1. If it's stupid but works, it isn't stupid.
  2. Professional are predictable, the world is filled with dangerous amateurs.

In other words, however numerous and skilled the hounds, they never see the world as a fox.

Starting with a precept that criminals will cleverly hack and subvert a system, using the Man's technology against Him is a rather academic view of people whom generally aren't. Practicality rules. And using tech in any way would tend to give an edge to law enforcement, not the other way around. The creator of technology always knows it better than a user, even a smart one. The Man makes/funds tech, criminals will almost always just be users who found a crack in the wall.

And anyway, technology aside, as I understand it, according to Mitnick, the best hacks have always been via social engineering, not device exploitation, though it helps sometimes.

For reference on your subject, check out the original Stainless Steel Rat trilogy by Harry Harrison. The series explores the topic of future criminals.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .