Driverless cars are definitely the stuff of sci-fi movies. Think about this scenario for a second: a highway full of cars driven by artificial intelligence zooming along lanes at controlled speeds, all of them avoiding accidents by wirelessly communicating with other cars. Sounds awesome, right? Umm…
However, just like in sci-fi movies, things can go horribly, horribly wrong.
All of a sudden, one hacker “jailbreaks” his car, overrides its settings, and successfully makes the vehicle go faster than the others. The software in other nearby cars isn’t equipped to deal with the tampered settings of this rogue automobile. In the blink of an eye, calculations go wrong, glitches pop up, tires screech, and bumpers collide. Your utopian vision of A.I.-controlled cars safely flowing along roads and highways? Up in flames.
Is this a nightmare scenario waiting to happen? Are these new driverless cars prone to hacker attacks?
During last year’s Black Hat security conference in Las Vegas, iSec security consultants were able to start a Subaru Outback’s engine and unlock its doors through a cellular network. They said it took them roughly two hours to figure out a way to gain access to the vehicle through their laptop.
It’s pretty alarming, as this demonstration proves there is a way for hackers to gain control of self-driving vehicles. However, several antivirus manufacturers as well as other groups and companies are currently hard at work, developing countermeasures and methods to protect cars from hackers. Intel, who now owns antivirus shop McAfee, is using the software company’s expertise to develop airtight defenses for both software and hardware, in order to prevent hacker attacks. Meanwhile iSec is working closely with car manufacturers to make sure systems are secure and free from any vulnerabilities that could be exploited.
It’s one thing when devices like computers, phones, or tablets are hacked, but when you’re talking about a moving vehicle – something that could potentially endanger lives – that’s another matter entirely. Hopefully companies like Intel do come up with foolproof ways to make self-driving vehicles free from the dirty keyboard-typing fingers of hackers with bad intentions.
As with any brand new technology, it’s only a matter of time before they iron out the glitches. And when they do, we’ll all be looking forward to a future wherein riding a car is a safer and easier experience for everyone.
So perhaps that sci-fi movie dream won’t have a bleak ending after all.
I read somewhere that everything is hackable, I think the recent Sony hack and various Twitter account hacks showed that. Especially with something as big as a grid of self driving cars, everything is indeed hackable. I can see a new method of assassination propping up along with the hackable self driving car…
Everything and anything is hackable, correct. We can all do things to minimize the risk to ourselves, but you cant trust everyone to do the same unfortunately. There will be some incidents, no doubt about it. Just hope they are not too damaging when they happen.
The fact that people will actually want to hack these self-driving cars is scary. Hacking the cars will give the hacker quite a bit of control over the car I’m guessing. Things like re-routing the car will most likely be possible, and worse things, I imagine.
Yes, everything is indeed hackable, but there are both sided, black hat hackers and white hack, that means if there are more white hat hackers hacking could be stopped, although if it was the other way around that wouldn’t be so good. A lot of black hat hackers that get caught turn to white hats though, which is good.
Oh yeah! The first thing that came into my mind when I read the words ”self driving cars” and ”hackable” was assassination! So murdering world leaders would be way easier for some governments if self driving cars became really popular. I could really see that happening.
So with that in mind, and even tho I’m an average person… I’d still not go for it. I rather drive my own car.
While this is quite interesting, I don’t feel that the example provided poses anymore risk than the current terrible drivers we already have. Driving is not safe, no matter who is controlling the car. It is reassuring to see that wireless security is a concern though. Tech has a long history of shipping out without proper security features in place.
The primary issue will always be human error, even when it comes to driverless cars. The problem at that point will lie in errors in coding where the car reacts in unpredictable ways when presented with specific conditions rather than some poor sod falling asleep at the wheel.
I’m not too worried about conventional “hacking” of an individual car – I’m more worried about hacking the systems involved in controlling the underlying logic for the entire grid. But, these are problems we already face – if a hacker can compromise the traffic light system for a city, it could cause mayhem.
From what I know, IRL hacking cars is almost impossible up to now, since any kind of “communication” their systems allow is restricted in an area really-really near them. The hack mentioned in the article was a PoC, “to show it’s possible”. Since there aren’t (yet) on the streets any cars with nav-driving systems connected to the Internet itself, we don’t have a problem. When they do reach the streets, though, and their systems connectivity isn’t anymore restricted to close-range NFC-like technologies… Well, then, yes, we’d better start worrying 🙂
Even if the vehicles themselves are 100% secure (which they’ll never be), the major point of attack would be the servers or what have you for the underlying control system. Whatever central point is controlling the routing and navigation of the vehicles is the weak point in any distributed model like this, and I’m sure it will run on relatively conventional computer systems.
Being able to bring a city’s transportation to its knees can be a major threat to public safety. What can you do when all the vehicles are unable to move or navigate properly, blocking public services and utilities from operating properly?
Now, that is science fiction – and will be for the next few years. Not that what you saying won’t ever be possible, but the first “self-driving cars” and almost any car with some kind of “autonomous systems” doesn’t, actually, rely on some remote server for its basic functionality – speed control, brakes, steering. That would be disastrous if, say, there was some lag in the connection – the car could “receive” a “signal to brake” to late to actually do it. So… Nope!
Todays (and tomorrows) models “pull” maps from “the cloud” (some internet service) and base their navigation on them, but actually navigate our world by using onboard sensors – Kinect-like cameras, lasers, whatever they creators thought was best for their solution.
So, nope, you can’t and won’t be able for a while to “hack a central server and push hundreds of cars off the grid”, creating havoc.
Inevitably. Every electronic out there is “hackable” in the sense that it can be modified by writing custom firmware assuming you can gain access physically to the circuitry.
This becomes doubly true if there’s any sort of DRM involved; installing DRM is almost inviting the modding community to challenge it. And where the modding community can squeeze out convenience and/or efficiency, they’ll certainly attempt to do so.
This sounds really, really scary. I don’t much like the idea of self-operating cars anyway, because forget hackers – what about the fact that AI will never be able to reach the same intelligence as a human driver has?
But yeah, I can definitely see how hacking would be a pretty big problem as well. They’d have to make the driving system operate “offline” first of all. It’s a tricky, tricky thing.
But also what if things went wrong on their own? No hacker or artificial intelligence involved? Things can go wrong with electronic devices, specially something as big as a self driving car. Bad things happen, they always find the way to happen…
So I could imagine the company being sued if something unexpected happened, like for example if the car didn’t stop when it had to and so on. I think a lot companies are so aware of this, so I doubt they’ll want to introduce this anytime soon.
This was an interesting article for me to come across. I hate it because everything now a days is hack able and there really is no way to stop it. Apps that I have had on my iPhone have been hacked multiple times then they had to shut the app down. I don’t trust people and I would never trust anyone if I bought this car that it would not get hacked.
Why are people so keen on the idea of self-driving cars? I personally think they could be a terrible idea! Hacking, malfunction, deliberate tampering etc… I mean I know this sort of thing still happens with cars NOW, but do we really want more ways for things to go wrong?
Have you seen Minority Report? “That’s why” 🙂
When you have to get from Point A to Point B, why should you also have to “do something” if you have the alternative to “not do something”, investing your time in something else apart from driving? When you go somewhere by using public transportation or a cab, don’t you sometimes read a book, play games or check your tasks on your smartphone, watch a vid on your tablet? Why not being able to do those in the privacy of your own “car” as well?
Plus, theoretically, apart from “that hacking problem”, autonomous cars could actually end being safer. You see, mistakes are a human privilege, as is loss of attention. How many people have died ’cause they were dumb enough to text while driving*? Well, if the car “drove itself”, they wouldn’t have met their end this way, would they?
* I don’t want to be disrespectful to anyone who lost a friend or relative under such circumstances, but I think that nobody disagrees (and if he does, he shouldn’t) that texting while driving IS a dumb thing to do.
Well, I have seemed to have thought they cars were amazing, but then you opened my eyes good sir.
Not only could a very experienced hacker do this, it could not only put your life at risk, but everybody’s. Imagine a hacker turning your car to block 2 lanes on the highway unexpectedly, so many cars could crash. Then he could set some thing so your engine over-heats and the car explodes, personally I would NEVER feel safe owning these cars.
Imagine what a Denial of Service attack could do. It could turn your engine or control switches off, leaving you with an out of control vehicle. A Denial of Service attack is so easy to do, kids that are even 8 years old can take down websites. 8 year olds committing these crimes may not know of the dangers, and could think it is a harmless prank to their friend’s parents, when all goes wrong.
Good thing you posted this, david. I feel as I should notify my anti viruses to make anti-hacker protection.
Of course they are. Especially with Google’s affinity for open source software.
Granted I’m sure they won’t release that kind of stuff to the general public, but we all know how easy it will be to get it eventually. However I highly doubt anyone would use them to do anything malicious.
Is it possible? Sure. But the type of people that CAN do it probably WOULDN’T. I know people get scared or whatever because of “hackers” and things like “Anonymous” but realistically it is just because they don’t understand it. Tech people will eventually take over the world (pretty much have already) but we will show everyone that we try to make the world a better place, not a bad one.
Yes, I was thinking about this concept the other day, what with the whole hacking story and news that Google has been developing cars.
I guess it really is possible, assuming there is no way that you can jam hacking efforts. We’ll just have to see how the technology develops. But I’d actually be more concerned about hacking efforts for planes and trains, which could prove to be a bigger story.
It’s only natural for this to happen. Hacking is just the process of finding innovative workarounds for features not intended in a program. So, a self-driving car will be hackable. And this will introduce a host of security issues that car companies will have to deal with, but I would assume that the most widespread hacks would not be malicious in nature, they’d be simply be enthusiast car owners modifying their car.
But if the most important controllers in a car are designed securely and if proper regulation is in place for what kind of modifications will be street-legal, I think this will actually result in the reduction of dangerous cars on the road. In my opinion, all the problems with self-driving cars are vastly outnumbered by the ones we already face with normal cars today.
Anything and everything that you can connect to in any way, or communicate with, is in fact hackable. However like everything else there are a whole lot of ways to protect things from getting hacked e.g adding better security and monitoring the cars every now and then.
I can see a few ways in which computer criminals could exploit this system in. One scenario would be that Black Hats now have the ability to threaten specific individuals for money or else they’ll hijack the car, or something along those lines.
With everything said, let’s hope for a bright future.
For every innovation, you have to assume many people will try to sway its purpose into a more nefarious one. And everything that is getting protected can be hacked at some point; we should be aware and being a bit paranoid is not always bad, it’s necessary!
The thought of these self-driving cars is terrifying without adding hacking into the equation! I have a mental image of articulated lorries going crazy on the motorway after a terrorist group have taken control!
Of course, I am simply scaremongering though! Driving will never be 100% safe and devices will never be 100 % secure. It would be a great excuse for being late into work though – “Sorry I’m late Boss, someone hacked into my car!”.
I never thought about this before but now that it has been mentioned, I’m sure someone, eventually, will find a way to hack a self-driving car. This could be used in future terrorists attacks and it would be pretty difficult to find the person behind it too. So now I have to wonder – do we even need self-driving cars?
I’m not sure I’m prepared to trust a self driven car. I see tech fail to do what it should on a daily basis and while its unavoidable that I put my life into the hands of some technology, I don’t think I’m quite ready to hand it over to an A.I. driver.
I seriously doubt that self-driving cars will go public any time soon. In 15 years at least… so I highly doubt that there will still be any obvious bugs left. Though, that is scary as heck. Let’s just all hope that when these cars eventually do go public, we’ll be at a state of ‘peace’ in the world where everyone understands the worth of human life… That will never happen. xD
Of course they’re hackable. Any computer is hackable and self-driving cars are no exception. A lot of things that were claimed to be ‘invulnerable’ have been hacked, like the iCloud servers and the recent The Interview controversy. And, no matter what, I believe that self-driving cars are more unsafe than safe.
I think one of the risks would be the provisions of governmental cooperation. I imagine that a target for the government when riding in a car of these is an opportunity that can not be missed. In this case the authorities may have methods to access the car. It isn’t crazy to think that government agencies demand that the software systems of such cars have a back door for them to control in these cases. If hackers take over these backdoors, problems are served.
This topic might be a little confusing to the older generation, but they will also be effected by this. Self-driving cars are definitely going to be a part of the future, so I believe companies should begin making them hacker-proof. I would pay more for more security.
We can learn a lot from self-driving cars. For me a self-driving car means a more laid-back life. If I were to include a self-driving car in one of my stories, I would be careful not to suggest any ideas about hacking a self-driving car. Furthermore, do people need a cyberliability insurance to cover hackable things?
I think I read almost everything is hackable nowadays, so with that in mind… No, I’d never consider getting one of those cars. Yes, even if parking it was way easier than parking an average car, I’d still not go for it. I think this kind of technology can be risky and we are definitely not ready for it, after all there are many malicious people out there… I don’t like the idea of me being at their mercy.
Unless there’s some ridiculous security measures, I do believe that smart cars are hackable, and I think it’s a good thing. There’s nothing wrong with wanting to get a little bit more out of a product you purchase.