Maybe it was IBM's Watson kicking human butt on Jeopardy. Maybe it's the inability of Japanese engineers to get runaway nuclear reactors under control. Maybe it's a hangover from Toyota's (TM) recall of millions of cars. For whatever reason, people are suddenly nervous about the security of the computer systems in their vehicles.
Their concern is totally overblown (and out of proportion with bigger threats). But it doesn't matter, because now the automakers are going to have to overcompensate. The vulnerability of critical stuff like brakes and engines came under scrutiny about this time last year, when a pair of academics proved they could hack the average family sedan. But worries have now intensified.
It's the network, stupid
Cars have been massively computerized for a long time. You'll remember that much of the speculation about Toyota's problems with sudden unintended acceleration was fueled by concerns about electronic throttle control and the lack of a physical connection between engine and driver in modern vehicles.
But over the next decade, cars are going to go well beyond the considerable (and very safe) computer assist they currently get. By 2020, cars will be thoroughly networked. You may not know where you are, but your car will. And it will know where all the other cars are. Because everything on the road will be communicating with everything else on the road.
How much trouble could this cause?
By and large, this will be a boon for safety, efficiency, and the elimination of traffic congestion. It will also look forward to a day when the driver will be taken out of the picture, as autonomous vehicles become prevalent. Of course, this kind of progress also means that people are going to nervous about someone of nefarious intent messing with the computing networks that control transportation.
It's not hard to imagine cloak-and-dagger scenarios in which engines and brakes are hacked to malfunction. In the wake of the many hacks that took place after the latest WikiLeaks data-dump, the general public now knows that even purportedly hyper-secure networks are actually vulnerable. This is tolerable when all that's affected are laptops and (briefly) bank websites. But when it comes to cars, safety isn't just required. It's assumed.
Fear of hacks will be worse that any actual hacking
There isn't much to be gained from hacking vehicle computers, except to sow chaos and fear (and also fraud, as certain hacks can manipulate odometer readouts). Still, the automakers are going to have to address this mounting concern by building layer upon layer of failsafes. At the moment, vehicle system are fairly discrete: When your car starts throwing error codes and setting off warning lights, the problem is confined to your vehicles.
Most people don't have code readers, and even those who do can't really act on the codes because they lack the technical and mechanical understanding to fix, say, a problem in the emission system. But as cars begin to push information to the computing cloud, the automakers will need to ensure that this deluge of networked information will be protected.
It will be expensive, and carmakers will have to eat it
Consumers want their vehicles to be more networked -- this is the dominant trend right now in the industry. But they don't want to necessarily pay hugely for the privilege. It's the same situation that automakers used to be in with safety. More safety costs more money, but once everybody understands how effective airbags can be in an accident, everybody starts demanding them.
In coming years, consumers simply won't buy an un-networked car (and actually, it may be difficult to operate a vehicle that doesn't contain basic self-awareness software). So the automakers won't have a choice about making the networked systems they include in cars secure to a fault, no matter what the cost.