Hackers taking control is biggest fear for self-driving cars

Telegraph

6 days ago

The public's primary concern over self-driving cars is that cyber hackers could commandeer the vehicles, new research suggests.
Almost seven out of 10 motorists believe self-driving vehicles will be 'highly susceptible' to hackers.
The survey comes after defence companies warned staff not to plug their smartphones into Chinese-made electric vehicles and the Ministry of Defence banned personnel from parking such cars within two miles of sensitive buildings.
Fears are rife that hackers answering to the Chinese government could use secret flaws in electric vehicle software to gain access to secrets or to eavesdrop on conversations inside the vehicles.
Yet those concerns may pale in comparison to a simpler and more direct fear: hackers taking control of cars away from motorists.
In a survey of 2,000 British adults, American tech company Cisco Systems found that the public's greatest fear over future self-driving cars is that hackers could break into them.
Seven in 10 motorists also believe that self-driving vehicles will be more dangerous than human-driven cars.
Half think that self-driving vehicles on British roads could cause a rise in traffic accidents, even though the technology is already approved for public use in limited circumstances.
Cisco also found that almost two-thirds, or 62 per cent, of drivers were worried about technology failures in self-driving cars.
'Trust and security are going to play a critical role in the future of self-driving vehicles. Automotive manufacturers must invest in appropriate network security architectures for connected vehicles to reinforce trust amongst users,' said Chintan Patel, Cisco's chief technology officer for Europe, the Middle East and Africa.
'With the rise of connected self-driving vehicles comes huge volumes of data, software and telemetry requirements,' he continued.
'Every Application Programming Interface (API) and point of connection in a vehicle can serve as an attack surface for bad actors.'
APIs are ways that computer programs expose internal data for further use. Smartphones that connect to a car's dashboard systems to control things such as music or the satnav do so through an API, for example.
Cyber security researchers have gone to extensive lengths to break into modern cars' onboard computers, as a way of highlighting potential safety risks.
One experiment in 2015 saw a Jeep Cherokee remotely turn off the vehicle's engine at 70mph, a flaw that the company acknowledged with a software update shortly afterwards.
Two years ago the Government approved Ford's BlueCruise self-driving system, which is permitted to take control of cars on English, Welsh and Scottish motorways at speeds of up to 70mph - although a human driver must always be behind the wheel and ready to step in.
The Government announced earlier this week that it had pushed the date for self-driving cars to be legal on British roads back to 2027.