Self-driving cars and civil liability for accidents
Many look to the advent of self-driving cars for the next revolution in the automotive industry.
The technology is revolutionary, but it comes with challenges. From the legal perspective, one of those challenges is how to attribute civil liability for loss and injury caused in accidents involving autonomous vehicles.
Accidents involving self-driving cars have occurred, primarily in the United States. There have been no accidents in Brazil so far, but it is prudent to contemplate how Brazilian courts might react when called upon to decide on liability and compensation for injuries and losses caused by self-driving cars.
The fundamental premise of self-driving cars is that their highly advanced artificial intelligence systems and sensors enable them to analyze their surroundings, make instantaneous decisions, and steer the vehicle with minimum human intervention, and so significantly reduce the incidence of accidents resulting from human failure.
Although the advantages are attractive, self-driving cars are not without risk. Technical issues, such as decision-making in complex and unforeseeable situations and interoperability between different systems, are factors that need to be considered. Infrastructure issues and the shared use of roads by self-driving and traditional vehicles can create challenging situations and, consequently, a higher incidence of accidents.
As a rule, liability for traffic accidents is fault-based, requiring proof that the driver was at fault in causing the accidents. In some situations, however, strict liability applies, which means that liability exists irrespective of fault. For example, while drivers' liability is fault-based, the owner of the vehicle has joint liability with the driver for the losses caused by the accident, regardless of whether the owner was at fault. Here the vehicle owner's liability is strict.
Strict liability can also apply under the Brazilian Consumer Defense Code (CDC), which provides that suppliers and manufacturers have strict liability for damage caused by their services and products. Automakers can fall under the scope of the CDC, and so have strict liability for losses caused by their vehicles.
When self-driving cars are involved, however, it will not be easy to determine who is responsible for accidents, since accidents may result from various causes. One of the first obstacles in determining liability is the difficulty of proving a failure in AI systems. Another problem is that accidents can be caused not by an identifiable failure in the system, but by a decision that the AI model accepted as correct.
As self-driving cars become a reality, specialists from a wide variety of sectors – the automotive industry, information technology, law, and insurance, to name a few – will have to work together to consider potential scenarios, devise measures for mitigating risks and losses, and establish frameworks attribution of liability.
While autonomous vehicles represent a notable advance in mobility, adopting this new technology requires a meticulous analysis of the risks involved, with the creation of appropriate regulations and well-defined systems of civil liability. These measures are essential to avoid legal uncertainties and diverging decisions by the courts. Even with these measures, however, it is likely that each case will have to be dealt with individually.
