Google has proposed a potential solution to the risk of connected vehicles becoming the victims of malicious hacks.  In its view, connected vehicles may need to be just a bit less connected.

The chief executive of Waymo, Google's driverless vehicle programme, has told the Financial Times that its self-driving cars will remain disconnected from the internet for the majority of the time and will "communicate with the outside world only when they need to".  This will do away with the need for a continuous cloud connection to the car, which will in turn reduce the scope for hackers to gain access.

Other carmakers' previous experiences have highlighted the cyber security risks posed to connected vehicles, and in particular the danger of a hacker accessing a car through one connected channel and then taking control of critical systems.  In 2015, a Jeep was hacked through its connected radio and the hackers subsequently infiltrated its main functions.

Waymo's approach to the cyber security problem contrasts with that of other manufacturers.  BMW, for example, has said that its driverless cars will need to transmit data and communicate with other vehicles continuously.

Cyber security is rightly one of the key concerns preoccupying the makers of connected vehicles and all participants in the autotech value chain.  Based on these reports, it seems we may start to see diverging ways of tackling this issue.

Some will promote the "always-connected" model, with all of the safety and functionality benefits it will purportedly bring.  Others will argue that connected vehicles will be more secure if they are more selective in how and when they connect to the internet (and they will need to demonstrate that there is no negative impact on the vehicle's operations as a result).  Lawmakers may add a further dimension as upcoming cyber security regulation starts to make its presence felt.