Self-driving car technology is improving so quickly that experts believe that it will be mainstream in five years. However, that huge shift will inevitably open the door to hackers targeting autonomous vehicles with potentially dangerous results.
That’s one of the key takeaways from a presentation about the future of automobiles and cybersecurity during the RSA computer security conference in San Francisco on Monday. With more cars now loaded with wireless Internet connections, hackers have more ways break into car technology than ever before, according to auto research company Kelley Blue Book, which has issued a report about consumer attitudes about car hacking and autonomous vehicles.
Get Data Sheet, Fortune’s technology newsletter.
Currently, the public doesn’t consider car hacking to be a big problem, according to the research, which is based on a survey of over a thousand people. The researchers found that most consumers consider distracted motorists and drunk drivers as two of the biggest safety concerns on the road.
Fears of cars being hacked turned out to be a relatively minor worry. It came in second to to last in a list of ten concerns, just ahead of carjackings, said Karl Brauer, a Kelley Blue Book senior director of automotive industry insights.
The attitude makes sense considering the average car on the road is around eleven years old, he explained. Older vehicles don’t have Internet connectivity, which makes them invulnerable to hacking.
However, when it comes to the cars of the future, respondents changed their tune. When the researchers asked respondents whether they are worried that future automobiles may be easily hacked, 62% of respondents said ‘yes.’ Additionally, over half of the respondents believed that car hacking will become a moderate to serious issue in the future.
Concerns about hacking haven’t deterred people from wanting the latest technology in their vehicles, however. New technology is a “make or break” factor for many consumers when choosing an automobile with one in three people saying it was important.
Over half those polled said they would be willing to pay for extra insurance to cover any losses from a vehicle hack. People were equally divided over whether vehicle manufactures should provide the insurance, the research showed.
When it comes to self-driving cars, the Kelly Blue Book experts explained that there would be many legal wrinkles to iron out because of the difficulty of determining who is liable for accidents. Is it the manufacturer who made the self-driving car or its owner?
For more on Google’s self driving cars, check out our video:
Brauer gave the example of a Google (goog) autonomous car that recently hit a bus. In that case, a test driver and the self-driving car incorrectly assumed that a bus would slow down when near the car, which led to a vehicle hitting the bus at less than 2 miles per hour.
Brauer said he recently discussed the matter with Google, which told him that it expects insurance companies or law enforcement to eventually decide who is at fault in such cases. Until then, the matter will continue to be a subject of debate.
In any case, Brauer said cars of the future like Google’s experimental vehicles will eventually have so many cameras, sensors, and data recording capabilities that “it better not be your fault” if human error causes a crash. Car manufactures will be easily able to determine who is responsible by reviewing all the imbedded technology.
“Nothing is going to happen to a connected car that will be a mystery,” said Brauer.