Google takes self-driving cars offline to prevent hacking
Google has confirmed that it is to disconnect its fleet of self-driving cars from the internet unless absolutely necessary to prevent their technology falling prey to hackers, according to the chief executive of its driverless car program.
The dramatic move has been made possible by the fact that onboard computers can remain autonomous even when connections to cloud servers are down, enhancing both safety and cyber security at a stroke.
It is not possible to disconnect from the outside world entirely however as autonomous vehicles require the means to communicate with other traffic and infrastructure to furnish them with advance warning of potential dangers.
Speaking to the Financial Times John Krafcik explained: “Our cars will communicate with the outside world only when they truly need to, so there will not be a continuous line that is able to be hacked, going into the car.”
Driverless cars are particularly vulnerable to criminals as their components are all connected on a central system, giving hackers multiple routes to break into the system and assume control of key functions.
The decision highlights just how seriously Google is treating cyber security after rival car manufacturers Nissan and Fiat Chrysler fell victim to criminal hacks.