Google's AI is allowed to take the wheel, says US regulator

It's full speed ahead for the tech giant's pet project, as driverless cars have just swerved a significant legal barrier.

by Rebecca Smith
Last Updated: 23 Mar 2016

The UK government might be getting excited about its £20m investment into eight new driverless car projects, but as Google's latest news shows, it's still a lot further down the road in terms of progress. A US transport regulator has just said a robot could meet the legal definition of a driver, which is a big win for Google (well, Alphabet – you know who we mean).

The letter to Google from the US National Highway Traffic Safety Administration (NHTSA), which sets the regulations on America’s roads, may seem fairly innocuous – it agrees with the company’s proposed interpretations of Federal Motor Vehicle Safety Standards. But the ripples of such an acknowledgement go much further – prior to this, any car without a human driver wouldn’t be considered permitted on the roads. But technological advancements, demonstrated with aplomb by Google, have convinced the NHTSA to adjust its stance. It's also a nod to Google's lobbying nous; this isn't so much a technological issue after all, it's a legislative one.

‘If no human occupant of the vehicle can actually drive the vehicle, it is more reasonable to identify the driver as whatever (as opposed to whoever) is doing the driving,’ the NHTSA said. ‘In this instance, an item of motor vehicle equipment, the Self-Driving System, is actually driving the vehicle.’

Google’s still got a way to go in convincing the general public that cars will be safer when humans have no control – there’s still a reluctance to embrace artificial intelligence, even if some of the misgivings surrounding its risks may be exaggerated. The tech firm, along with others jostling to get ahead in the driverless car game, will probably need to do more in raising awareness and understanding of the safety of the systems used. Its AI system will make what could be incredibly serious decisions based on data from a range of sensors, maps and cameras. Driving is not the same as chess.

It'll likely be the transition period which causes the most difficulties – when driverless cars are being introduced on the roads, but the majority of vehicles are still driven by people. Autonomous cars will be able to predict each other's movements with ease; judging what a human driver is up to is another matter altogether.

Getting the NHTSA on board is a big step in gaining credibility though, particularly after last year’s hold up when the California Department of Motor Vehicles said that any autonomous vehicles being tested on the state’s roads must have a steering wheel and pedals.

The regulator’s agreement with Google did also state that it didn't necessarily resolve all of the issues raised by the company (in two letters sent in November and January). The NHTSA said Google would have to work around federal rules such as the need for brake pedals, potentially by the tech firm petitioning for an exemption as it carries on with its testing.

Google may be revving up its efforts; impatient to roll out its driverless vehicles as soon as possible, but the NHTSA did warn changing the definition of a ‘driver’ could take a while. In the meantime, Google might want to think about how to win over the general public and what else can be done to navigate other potential roadblocks on the horizon, including the questions surrounding fault and liability.

Find this article useful?

Get more great articles like this in your inbox every lunchtime