Driverless cars could be used for assassination, says Attorney General

“There is no Internet-connected system where you can build a wall that’s high enough or deep enough to keep a dedicated nation-state adversary or a sophisticated criminal group out of the system,” said Carlin. The prospect of an international hit team controlling a driverless car with a joystick, perhaps working a windowless room continents away, may sound like a spy fiction plot, but cannot be discounted.

An incident in 2015, when a Jeep Cherokee was hacked to demonstrate potential remote takeover, raised the security concerns of driver-controlled vehicles with onboard automotive system connectivity. That hacking also resulted in a recall of 1.5 million vehicles and questions from federal safety agencies about why Fiat Chrysler Automobiles NV didn’t report the known problem 18 months earlier.

With the international interest and investment in furthering autonomous car development, re-emphasis on security from outside threats is essential. When self-driving cars don’t have a driver to take over in case a vehicle starts speeding toward a bridge abutment or heading for a river, the onboard systems better be tight. Vehicle crossovers of GPS location systems, on-board cellular and Wi-Fi communications, 360-degree vehicle cameras, and links to automotive diagnostic and control systems raise warning flags about the hacking threat.

“If you were able to do something that could affect a large scale of an industry — like 100,000 cars — you could see that being in the arsenal of a nation-state’s tool kit as a new form of warfare,” Carlin said. “We’ve seen rogue nation states try to assassinate those that do not share their beliefs. If they were able to do it remotely through a car, I don’t see why they consider that a safe zone.”