How self-driving cars could become weapons of terror
sfchronicle.com
sfchronicle.com
Ever since at least last year I keep waiting for the news to break out: drone + computer vision + navigation system + some kind of weapon (maybe a gun, maybe a poison dart) - all that stuff sent to seek someone out and "deal" with them. Totally doable.
Just a matter of time.
Holy JC on a bicycle, I didn't even think of that.
Yeah, good point. Forgot about those.
Coat the explosive material with a (hemi)sphere of ball bearings, and you have a flying claymore mine.
The Finnish military has a nickname for large claymore mines: Refugee TV. I can't imagine what they would call a swarm of drone variants.
Car only needs a computer - not a network. Modern cars already have computers. Picture hackers employed by a yadda yadda remotely disabling power-steering on cars across the nation or disabling ABS and you have the same disaster scenario in an already-existing attack vector.
>Or envision a lone-wolf terrorist loading explosives into a car and programming it to drive to a targeted building or public space.
Already possible. Even Mythbusters sets up "life size" RC cars for some of their crash-related myths.
I really hate puff scare pieces.
It's already possible, but it takes a lot of work, and (AFAIK), you don't end up with a vehicle that can be programmed with destination coordinates tens or hundreds of kilometers away and then left to its own devices.
There's always the possibility of a lone-wolf, obsessed Unabomber-style terrorist who has their own machine shop and whatnot, but that's an edge case. The potential danger here (IMO) is the terrorist who can steal a few autonomous cars from parking lots (or call up a few autonomous taxis), load them up with cellphone-detonated explosives, and be all set. That scenario doesn't require anything really technically-challenging, and IMO is very likely to occur.
The random "crash a bunch of cars on the freeway" scenario is not as likely, IMO, because the number of people who are interested in killing tens or hundreds of random people is lower than the number who are interested in killing a few very specific people, or destroying a very specific building/piece of infrastructure.
I don't think this is a reason to disallow autonomous vehicles, but I don't have a good answer for how to address it, either.
How does that make sense? They are talking about "hackers" seizing millions of cars remotely in one paragraph, on the next we have perfect technology that can sniff dangerous chemicals and disable a car. But wait, aren't those people supposed to be hackers? Why can't they disable the sensors then?
Also, what's preventing them from retrofitting a car to be remote controlled, if they are so skilled? Newer cars are almost "fly-by-wire" anyway already.
Also, if we could do that might as well also have it call the police with the exact GPS location and drive the perp to a remote bomb disposal location.
Alternatives are cheaper, already available and you can't stop someone who really does want to spread terror.
That's not very comforting. We're just getting to the point where it's becoming mainstream for corporate IT to understand that "the firewall" is not some magic line of security demarcation where "evil" stays firmly on "its side".
"Firewalls are like seat belts, not magic force fields."
Seems like an apt analogy for this one.
Why does there even need to be a firewall. It should be air-gapped in the truest sense.
Not defending this behavior, but that's a reason..
But if you have more people ready to accept the guilt tripping and ideological shame it takes to make a radical, I'm guessing those capabilities will dry up. We really ought to be focusing on the intent rather than the means, because the means is obviously already common.
(Also "Demon Seed")