The Blame Game: Who pays if driverless cars crash?
Autonomous vehicles are coming to make traffic safer. But who is responsible, if “robot cars” do crash? Accident liability in a driverless future raises some tricky questions. We asked experts from both sides of the Atlantic about it.
Road traffic as we know it is on the verge of a sea change: the transition from vehicles driven by humans to vehicles driven more and more by themselves. Built to an unprecedented level of safety, these connected cars will dramatically lower the number of accidents worldwide.
But as automation is gradually introduced on our roads, there might still be crashes. Imagine a car – traveling in self-driving mode – hitting a pedestrian. Or two autonomous cars smashing into each other.
Scenarios like these are causing legal experts some headaches: who’s at fault if automation fails? And who pays, if robot systems cause a collision?
No changes to the law
Some commentators even go so far as to doubt whether automated driving has a future at all amidst unresolved regulatory and liability questions. But there is reason for optimism, says Jacob Fuest. “We don’t see liability getting more complicated in a driverless future,” the Head of the Automotive Innovation Center at Allianz tells 2025AD. Existing laws in a majority of countries in the European Union would well be able to handle liability even in complex crash scenarios. “Currently, we see no need for action on the side of insurance and liability law,” Fuest explains.
Across the pond, legal experts tend to agree. “Law is a flexible creature”, U.S. attorney Andrew Garza explains in an interview with 2025AD. According to Garza, who heads the Law Offices of Andrew P. Garza in Connecticut, future liability cases will likely be handled the same way as today. Even if they involve autonomous cars, they can be addressed through existing product liability regulations. “Will there be a learning curve? Sure,” Mr Garza says. “But ultimately, our liability framework that has existed for hundreds of years will be well-adapted to handle the change."
A human will still be held accountable
Now what does this mean in practice? Imagine you’re driving on a highway in the year 2025. You’re checking e-mail while your brand new car is cruising in autonomous mode. All of a sudden, it wrongfully changes the lane and touches an “innocent” vehicle. According to Jacob Fuest, it is you yourself, who will, in the first instance, be liable for the damage caused by your car. “The basic legal principle of strict liability will endure,” says Fuest. This means: the person in charge of the car is liable for both individual mistakes and defects of the vehicle.
So no matter if you or your car were in control at the moment of the collision, accountability will first be assigned to you. Your vehicle insurance will pay for damages done to other road users. As Mr Fuest explains, the overarching doctrine of victim protection makes a swift, no-questions-asked compensation an utter necessity. “We cannot and will never expect crash victims to press charges against manufacturers or other parties.”
But of course, that is not the end of it. What if the crash wasn’t really your fault? What if a malfunction of the autonomous features caused the accident? In this case, your insurance will, in a second step, approach the car’s manufacturer or other parties involved to seek recovery for its advance payments. In the same way, the insurer will proceed with damages to your own car.
Who’s to blame? A black box will determine that
When insurances and manufacturers sit down, the real blame game begins. The crunch question will be: who was the perpetrator – man or machine? And if the human is ruled out, then what exactly caused the crash? A brake malfunction or faulty sensor information? Compromised software or flawed mapping data?
In connected vehicles that communicate with other cars and perceive their surrounding world in 3D, all of this seems possible. To unravel the complex technical entanglements of a driverless accident, data will be key. Specifically, so-called event data recorders may bring everybody closer to an answer.
Once automated driving arrives on our roads, these devices will become more prevalent. Fire and shock-proof, the appliances are similar to the black boxes of airplanes. Working like a log book, they store relevant information to shed light on what went on in the decisive seconds before a crash. In future automated driving, they will be designed to also record the car’s driving mode (autonomous or manual). The system will be able to register faults in the assistance systems, software bugs or network failures that may freeze a vehicle’s communication with other cars (V2X).
“This will resolve most liability issues,” Andrew Garza predicts. “Maybe there won’t even be a need for eye-witnesses to say ‘this is what I saw’. Instead, we’ll let the data show. That’s the best evidence we can have. Reviewing that data will be a game-changer.”
Open questions remain
While the data from these recorders may provide answers, their collection raises further questions. According to Jacob Fuest, key challenges will be to solve sensitive issues of data privacy and to standardize processes between carmakers and insurers: “Which exact data do we need to determine what happened? And how do we ask permission to access it? First and foremost, any given data is owned by the customer.”
With an increasing degree of automation, Jacob Fuest expects the onus to be on manufacturers to prove their cars were not at fault in the event of a crash. Volvo has already claimed it will take full liability for damages caused by its driverless cars. Once carmakers start selling more automated vehicles and we start living in an autonomous car reality, “automakers will be taken by their word”, Andrew Garza expects.
Things will get trickier on the manufacturer’s side as more and more external parties such as software developers or mobile network operators enter the field. “With automated driving on the rise, we have a significant increase of players,” Fuest observes. In order to address liability proactively, there will likely be more mutual agreements on indemnification between parties to protect themselves against damage claims.
The sword of Damocles
In the U.S., punitive damages may still pose the biggest threat to the car industry. After all, if an autonomous vehicle is “found guilty”, carmakers might be confronted with further claims regarding compensation for crash victims’ personal suffering.
Could the classic “million dollar trial”, hanging like a sword of Damocles over the industry, impede or even prevent consumer access to automated mobility? Chances are it won’t. “I do think autonomous driving is an inevitability,” Mr Garza asserts. “These cars will be available. And although it’s also inevitable that some carmakers will be too cautious, enough companies will recognize the benefits in the long haul.” Also, Jacob Fuest is certain: "Autonomous driving is a major step towards a future with less road accidents. That's why we work so closely with the automotive industry on removing potential obstacles on the way to a safer future."
What’s your view on future liability – how should it be handled when it comes to driverless cars? And which hurdles do you think will still need to be removed? Tell us in your comments below!