As we stand on the verge of a new era of driverless cars, we need to consider the significant challenges this presents for lawmakers and the legal and insurance professions. Importantly, we need to consider the ethical and legal challenges for the decades of transition, when autonomous and human-driven cars will operate side-by-side on the roads.
Google, Uber, Mercedes and others are working on driverless prototypes for mass production. Parliament is preparing for a world of autonomous vehicles - most recently through publishing the Vehicle Technology and Aviation Bill in March this year. While some pundits anticipate the first fleet to be expensive toys for wealthy trendsetters (raising the possible return of the car-related class frictions that upended British society in the early 20th century), others have predicted that autonomy will deliver savings - a recent study in the US predicted that without the costs of a driver, autonomous taxis could run on as little as 35 cents per mile.
Whether expensive or cheap, there will be a long period of transition from human driving to full automation. People who have recently bought a traditional car will want to get proper use of it for several years to come, before investing in a new automated vehicle. And some people may choose never to switch, preferring the sense of control and freedom that being a passenger simply does not offer. As human and computer drivers share the roads, challenges will arise. Given that human error is often to blame for road accidents, the burden of suspicion after a collision will fall on the human driver. People who believe that an automated vehicle was at fault in an accident may face an uphill legal battle against powerful manufacturers to prove their case.
The insurance challenges in this transition period will be significant. In the UK, individual drivers are insured. In France, insurance covers the vehicle, not the driver. Given that users of automated vehicles will essentially be passengers, it is hard to see how culpability for an accident could rest with them. The owners rather than the users may be liable, but it seems more likely that insurance will cover the manufacturer or the product. In a hybrid world, collisions could, therefore, pit drivers' insurance companies against manufacturers' insurance firms. While a lengthy legal case between insurers and manufacturers proceeds, injured parties will need some form of immediate assistance (say, for medical treatment or lost earnings), to avoid satellite litigation while the main claim is considered by the courts.
A Bank of England paper in March this year considered some of these thorny questions and warned that the insurance industry would need to transform to deal with an estimated 80 percent automated vehicle fleet in the UK by 2040. The Bank estimated a contraction in insurance premium revenues of at least 21 percent, and perhaps significantly more. It remains unclear how the insurance sector will react to these declining revenues in an age where the fault is much more difficult to determine. Parties at fault are no longer restricted to the drivers involved in a collision. Manufacturers, fleet owners, overseas tech firms, satellite navigation systems and other parties all could be drawn into a complex, lengthy and expensive process after each accident. While the number of accidents may decline, the issues raised each time they occur will be much more complex.