Until recently, driverless cars felt like something from science fiction. Today, they are moving steadily towards becoming a reality on UK roads, with pilot schemes already taking place in locations such as Milton Keynes. While this promises significant benefits in terms of road safety, fuel economy and more efficient travel, it also raises fundamental questions for road traffic laws that have evolved over almost two centuries, with much of the current legislative framework dating from the 1980s. As vehicles become increasingly capable of driving themselves, difficult questions arise over who is legally responsible when something goes wrong and whether liability should rest with the user, the manufacturer or the software developer.
What is the current legal position?
Our road traffic laws have always assumed there is a driver engaging in the voluntary act of driving the vehicle. This is why, in exceptional circumstances, it can be a defence following an accident to show there was no voluntary act, for example because the driver suffered an unforeseen blackout or lost control after being unexpectedly attacked by a swarm of bees. Driverless cars challenge those assumptions in a way Parliament has never previously had to address.
Autonomous vehicles are not currently legal on UK roads. However, many of us will be aware of features within cars that have some autonomous elements, for example lane assist, radar adaptive cruise control, automatic emergency braking, parking assist, road sign recognition and so on. Current technology ranges from systems that assist the driver through to fully autonomous vehicles requiring no human input. This represents the long-term ambition for autonomous vehicles, where no human driver is required and the traditional concepts of holding a driving licence or being fit to drive may become largely redundant.
Trials of highly automated vehicles have taken place in Bristol, Coventry, Greenwich and Milton Keynes, but currently only vehicles up to Level 2 automation are legal on UK roads. However, with the Automated Vehicles Act 2024 due to come into force in late 2027, all that is set to change.
Who is criminally liable?
The current legislation means that even where the vehicle is performing an automated driving function, the person behind the wheel is still deemed to be the ‘driver’ and is therefore liable for any offences committed during that time.
They may, however, have a special reasons argument, which can be used to persuade the court not to impose the usual penalty points or driving ban because, although technically guilty, it would be unfair to impose the usual penalty.
In such scenarios, the courts are likely to distinguish between blind reliance on technology and reasonable reliance on an approved autonomous system. Because special reasons arguments are notoriously fact-sensitive, two seemingly similar cases could produce very different outcomes. This could be because different types of roads, for example a motorway compared with a winding country road, make reliance on driving assistance more or less reasonable, or because of the warnings, if any, the vehicle gave to alert the driver to the need to take action.
How will the law change?
The definitions of ‘driver’, ‘user’ and ‘owner’ which feature in our existing road traffic laws are well established, but section 46 of the Automated Vehicles Act 2024 creates a new category of being a ‘user-in-charge’. The Act is designed to remove criminal liability when the car is in fully automated mode. It also provides a number of exemptions to this immunity from prosecution, including where the vehicle ‘demands’ that the user-in-charge takes over the driving.
Complications are likely to arise in establishing which mode the vehicle was in at the time of the offence or when it was transitioning between the two. Other considerations are whether a user-in-charge is required to intervene where the vehicle in automated mode appears to be malfunctioning.
Liability will likely shift towards manufacturers and/or software developers for offences committed whilst in fully automated mode, but this does not mean that a user-in-charge could never be prosecuted. They may still commit offences by ignoring repeated takeover requests, placing the vehicle into autonomous mode where it is not authorised, deliberately failing to intervene or tampering with the vehicle’s systems.
It is important to distinguish between the driver assistance systems with which we are already familiar and authorised self-driving features certified under the Automated Vehicles Act. Many motorists assume that features such as adaptive cruise control already make their vehicle “self-driving”, whereas legally these are very different concepts.
How will investigations change?
Proving which mode the vehicle was actually in may become the modern equivalent of proving who was driving and manufacturers are likely to hold enormous quantities of electronic evidence in this regard. Beyond just expertise in accident investigations, as is currently relied on, police officers may need specialist training simply to identify whether the vehicle was in autonomous mode.
Unlike conventional motoring prosecutions, key evidence may not be in the possession of either the prosecution or the defendant, but rather the vehicle manufacturer or software provider. Ensuring that this material is preserved and disclosed is likely to become an important aspect of future cases.
CCTV and dashcam footage, so heavily relied on currently, may no longer tell the whole story. Determining liability could depend upon internal vehicle logs recording steering inputs, software decisions and precisely when control passed between the vehicle and the user.
Therefore police investigators may increasingly find themselves needing to obtain evidence from vehicle manufacturers rather than relying solely upon traditional collision investigation techniques. Motoring solicitors will similarly need to familiarise themselves with this new kind of ‘technical’ evidence.
The Future of Road Traffic Law
As vehicles become increasingly capable of making driving decisions themselves, criminal investigations are likely to become less about eyewitness evidence and more about software logs, electronic data and establishing precisely when legal responsibility passed between the vehicle and the person behind the wheel.
Those questions are unlikely to have straightforward answers. Investigators, prosecutors and defence lawyers will increasingly need to understand vehicle telemetry, manufacturer data and software evidence when determining whether autonomous mode was lawfully engaged and where criminal liability rests.
One thing appears certain: the early cases brought under the new legislation are likely to generate significant litigation and shape the future development of road traffic law for many years to come.
About the Author
Clare Galo is a Senior Solicitor at Reeds Solicitors, specialising in motoring law and criminal defence. Having qualified in 2004, she has spent over two decades representing clients in police stations and magistrates’ courts, with particular expertise in road traffic offences, including drink driving, serious injury by careless driving, special reasons arguments and exceptional hardship applications. She has a particular interest in the legal challenges presented by emerging vehicle technology and the future development of road traffic law.
To speak to Clare or a member of our specialist motoring team, call 0333 240 7373 or email [email protected]
