Civil Machines: The Current State of Autonomous Vehicle Liability – Rail, Road & Cycling

In 2016, automakers predicted that autonomous vehicles (“AVs”) would hit the road for widespread consumer use as early as 2021. It’s now 2022 and human-operated vehicles still dominate the road. While the reasons for the delay aren’t particularly clear, it’s obvious that VAs still have a long way to go (pun intended). A recent study by the American Automobile Association found that self-driving vehicles attempting to negotiate an intersection with a cyclist hit the cyclist 33% of the time. (https://newsroom.aaa.com/wp-content/uploads/2022/05/E-1_Research-Report_2021-ADA-Evaluation_FINAL_4-13-22.pdf)

Despite their slow start, the emergence of AVs on major highways and urban roads is likely inevitable. The advent of AVs means that human drivers will play an ever smaller role in vehicle operation. VAs are unlikely to be completely autonomous from their inception; rather, the initial phases of AVs will be semi-autonomous, allowing drivers to hand over control of their vehicle to artificial intelligence or automated driving systems (“ADS”) while their vehicle is in motion.

In theory, AVs will and should reduce the number of accidents we have on our roads. However, accidents are unlikely to be a thing of the past – especially they will be as long as humans retain some control over the operation of AVs. Even if human error is completely erased from the equation, it is possible that future accidents could result from faulty artificial intelligence, software or product design. As responsibility for future accidents shifts from drivers to manufacturers, the law will also have to evolve with them. However, there has been little discussion about how we legislate and govern these vehicles once they are ready to roll. One area of ​​concern that will need to be addressed will be the determination and allocation of civil liability for bodily injury resulting from motor vehicle accidents involving AVs. Fortunately for us, our friends across the Atlantic have taken over in this regard.

Automated and Electric Vehicles Act 2018

In the UK, the Automated and Electric Vehicles Act 2018 (“AEVA”) received Royal Assent on July 19, 2018. (https://www.legislation.gov.uk/ukpga/2018/18/contents /enacted) AEVA takes a radically different approach to the current traditional insurance regime in that it mandates direct liability of insurers for accidents caused by AVs, while the vehicle is not under the immediate control of a human driver. In other words, the insurer is liable for all damages caused by the AV, whether these damages are suffered by the insured or a third party.

In January 2022, the UK Law Commission completed a joint report on the regulatory framework for AV vehicles in the UK. (https://www.lawcom.gov.uk/project/automated-vehicles/) This report has provided an overview of the potential issues around third party liability, among other issues, that are likely to arise as autonomous vehicles become more popular. ubiquitous in Canada.

Problem #1: Exclusion from Coverage

Section 3(2) of the AEVA excludes coverage in cases where the accident is “wholly due to the negligence of the person in allowing the vehicle to commence driving itself while not appropriate to do so”. This section places liability on the driver if he negligently engages the ADS while operating the AV.

Some UK stakeholders were concerned that Article 3(2) places a heavy burden on individual drivers to discern whether to engage ADS while using AV. For example, the terms of use of could state that the ADS should not be engaged during heavy rain or snowfall, leaving many drivers to use their own weather skills to discern the suitability of the weather before engaging the ADS. ‘ADS.

The UK Law Commission has suggested that a possible solution to this dilemma is to ensure that AVs can only operate under ideal conditions. But limiting ADS to only the most ideal conditions will likely deter consumers in countries like Canada from embracing the technology all together. Therefore, such an exclusion may not be feasible if the initial phases of the technology are to be deployed in Canada.

Problem #2: Secondary Claims

The AEVA allows insurers directly liable vis-à-vis an injured person to take additional action against any other party that may be at fault, whether the driver of another conventional vehicle and/or manufacturer.

Section 5 of the AEVA provides that a secondary claim against another offending party may only be brought after the aggrieved party’s claim is settled either by judgment, arbitration award or binding settlement.

It seems the intent of Section 5 is to incentivize early resolution of primary claims since secondary claims against responsible drivers or automakers could be stuck in court for years.

Issue #3: Causation and Contributory Negligence

Section 2 of the AEVA states that coverage will apply when the accident is “caused” by the AV while ADS is activated. It is not entirely clear from AEVA whether coverage applies if the VA only does causes the accident or whether there must be an element of defaultfrom the VA before coverage becomes applicable.

The UK Law Commission has taken a rather evasive stance on causation, citing that it should be left to the courts to decide whether causation requires fault for AEVA to apply.

Presumably, UK courts will follow the common law that fault is a necessary element of a person’s liability for damages. Deviating from current case law would greatly increase the likelihood of claims and could make insurance coverage too expensive for insurers to purchase and too expensive for consumers to purchase.

Additionally, Section 3(1) of the AEVA provides that contributory negligence will remain in actions involving VAs. As such, the liability of AVs will be reduced to the extent of the negligence of the injured party.

Subsection 3(1) then specifies that contributory negligence “shall apply to a claim relating to the accident brought by the injured party against a person other than the insurer or the owner”. This article is quite vague and requires some legal gymnastics to be understood because it creates two counterfactual situations:

  1. The court must treat the claim as if it had been brought against a person other than the insurer or the owner of the vehicle under tort law, and

  2. That the insurer is at fault because of the behavior of the AV.

Alex Glassbrook of UK firm Temple Garden Chambers LLP, suggested that one possible interpretation of Article 3(1) is that it “provides for a fictitious human comparator in place of the car, to avoid difficult comparisons between reasoning human and machine reasoning when evaluating comparative fault.” (https://aevlaw.com/blog/overview-of-the-automated-and-electric-vehicles-act-2018/)

Although some UK stakeholders have expressed concerns about the ambiguity of Article 3(1), the UK Law Commission has indicated that the AEVA is “good enough for now”, citing that the courts are well equipped to investigate any issues that may arise.

Conclusion

It makes perfect sense for Canadian common law jurisdictions to adopt a legislative scheme that closely resembles the EIA given the similarities in our respective legal systems. Of course, the Canadian scheme will likely be modified to account for regional differences (eg, climate) between nations. But the UK’s experience will be a valuable ‘guinea pig’ to study as we prepare our own legislative framework for the brave new world of AVs.

The content of this article is intended to provide a general guide on the subject. Specialist advice should be sought regarding your particular situation.