Autonomous cars will and to some extent have become a reality in the not so distant future. With companies such as Tesla leading the way and showing consumers what they might expect regarding the consumer use of autonomous vehicles, a lot of legal questions arise. One of the biggest issues that developers of autonomous cars are faced with is liability. Uncertainty in this area, which results in the fear of having to pay high damages, cripples innovation.1 With this in mind the European Parliament has ordered an Added Value Assessment, which would accompany an initiative legislative report. One of the subjects researched in this assessment is the applicability of the Product Liability Directive (PLD) on the evolving field of autonomous cars and other vehicles. What happens when the PLD is applied to cases involving autonomous cars and where might problems arise?
An important feature of the PLD is that it implicates a strict liability. Meaning that the producer shall be held liable for damages caused by a defect in the product even in the absence of any fault or negligence on the producer’s part or of a breach of contract with the claimant. Because of this and other legal implications imposed by the Directive, it is a piece of legislation that producers must most certainly keep in their minds when assessing their financial and corporate risks.
The definition of ‘product’ under the PLD
The main rule of the PLD is stated in Article 1:
“The producer shall be liable for damage caused by a defect in his product.”
From this main rule we can deduce three main points of concern when investigating the application of the PLD to crashes involving autonomous cars. First, there is the definition of ‘product’. Second, there is the notion of defect under the PLD. And third, there is a question regarding the type of damages that are covered under the PLD and to whom this applies.
As stated by the European Added Value Unit in its report about the value of the PLD in an autonomous vehicle context, the text of the PLD was mainly written to deal with damages related to hardware failure. However, since a great deal of the driving task is, and will only grow to be, performed by software, does the current notion of ‘product’ still suffice? The main hurdle to overcome when regarding software as a ‘product’ in the meaning of the PLD, appears to be the fact that it isn’t a moveable object. It lacks tangibility. Some have argued that this shouldn’t be a problem, as software is always stored on a tangible object, and therefore it fulfils the requirement of tangibility. Another argument for the applicability of the PLD to software is the fact that electricity is also expressly included in the scope of the PLD. Others argue however that the fact that electricity is expressly mentioned means that any other immovable object is, therefore, not within the scope of the PLD, following the unius inclusion est alterius exclusion-rule. Arguing that, using a teleological interpretation, software falls within the scope of the PLD.
Defectiveness under the PLD
One of the biggest challenges to overcome for an injured person is proving a causal link between the defective product and the damage it has caused. With the growing complexity of autonomous vehicles this will prove to be more and more difficult. One has to prove that the autonomous car does not provide the safety that a person is entitled to expect. The question regarding this ‘legitimate expectation’ is what the general public believes the safety expectations of a certain product are. This is an important aspect of the notion of defectiveness, as it is something manufacturers have influence on. By reminding drivers and the overall general public that the car still needs constant supervision by the human driver, the manufacturer can avoid liability as it will then be the ‘legitimate expectation’ of the public safety that the car is not yet able to provide an autonomous driving experience that is fully safe. However, most people do expect that an autonomous car is able to deliver such an experience. Therefore, there is most certainly a giant liability risk for the producer in this field if they do not, in some way, manage to ease the public expectation.
There is however not only a massive financial risk for the producer but also on the injured person, as scientifically unknown risks will be placed upon the injured person according to Article 7 of the PLD. Due to the highly technical nature of the notion of defect proving fault could likely prove to be an undoable challenge for consumers. In such a way, the PLD will fail to provide the financial safety it aims to deliver.
Damages covered by the PLD
The final point of contention is that, in its current form, only private consumers are protected by the PLD, leaving professional entities out in the cold. This is the case due to the fact that one of the main goals of the PLD is to provide consumer protection, meaning professionals do not fall within the scope of the Directive. Should this change with an eye on the technological possibilities and the growing amount of professional use of the autonomous car?
Further, as the Added Value Assessment states, the PLD does not cover damage to the autonomous vehicle itself and it is limited with regard to property damage. This means that the party that has occurred damage has to ‘fall back’ on national tort law in order to get these damages covered, which means that one still has to prove fault at the side of the producer, the one thing the PLD tries to avoid by imposing a strict liability.
The technological advancements that have been made in the field of autonomous vehicles have caused for significant legal problems regarding the liability of producers and users under the PLD. The notions of ‘product’ and ‘defect’ used in the PLD are somewhat outdated and in need of revision if the PLD is to be applied to such use cases in the future. It can be applauded that the EU has started research into the subject to overcome a major regulatory delay when the autonomous technology advances past the current regulation.