Sunday, 3 February 2019
Answer to Consultation Question 18
Is there a need to review the way in which product liability under the Consumer Protection Act 1987 applies to defective software installed into automated vehicles?
I infer from the Law Commission’s limited terms of reference, that this question is primarily concerned with secondary claims brought by motor insurers under s5 AEVA 2108; not the primary accident victims whom the entire edifice of Part VI Road Traffic Act 1988 and S2 AEVA 2018 is supposed to benefit. In which case, strictly speaking this question only concerns those vehicles likely to be listed under s1, i.e. high or full automation – (see LCCR 240 para 2.55 and footnote 83).
Yet the Law Commission expressly concedes the need to widen the scope of its consultation (see para 1.14) and this much is clear from some of its questions, such as Q6, 7,8 and 12.
The Consumer Protection Act 1987 (CPA) fully implements the Directive 85/374/EEC on Product Liability and can be easily revised after Brexit should need arise. Two important concerns, raised in para 3.9 of the Government Response to its consultation on Advanced Driver Assistance Systems and Automated, from January 2017, are that this legislation might be inadequate for insurers wishing to invoke their statutory right of recoupment, because (i) product liability is optional, and (ii) no statutory minimum amount of cover is set. Against that, it should be acknowledged that motor insurers are in an advantageous commercial position to limit their exposure to risk by negotiating effective reimbursement guarantees from manufacturers and suppliers, in return for offering competitive (and from the ultimate consumer’s perspective, affordable) premiums. Manufacturers who fail to incept a Volvo style compensatory guarantee are likely to find that the cost of insuring their vehicles might make their products unattractive to consumers. Problems will nevertheless arise where causation is disputed or needs to be apportioned. However, the motor insurance industry is best placed to confront these challenges as it has the necessary expertise and resources as well as a strategic overview that few private individuals could hope to command. It is conceivable that the MIB could co-ordinate collective actions against manufacturers and or develop a specialist expertise in these claims, as opposed to individual insurers. The government should ensure that its regulation of motor insurers ensures that the industry has taken all appropriate measures to safeguard it exposure to insolvency and to guarantee that victims recover their full compensatory entitlement. This is not something that can safely be left to the industry to self-regulate with complete autonomy.
If the strict liability provisions of s2 are to be confined to highly and fully automated vehicles then this discriminates against victims of what are arguably even more dangerous forms of automation that are already on our roads or which will shortly be introduced (SAE L2 & 3 automation).
The UK’s adversarial civil justice system, whilst striving to ensure fairness and to put individual parties on an equal footing on a case by case basis, has serious shortcomings when it comes to an unequal contest between private individuals and international corporations. Fixed recoverable costs and the proportionality rule shift the tactical advantage heavily to the advantage of a corporate opponent who can afford to adopt untenable or unreasonable stances in individual cases if it serves its wider commercial interests. Private individuals generally do not have the same expertise and usuallyt nothing approaching a parity of resources. Accidents caused by existing and close to market levels of vehicle automation (SAE L2 & 3) are likely to prove to be fiendishly technical, time-consuming and costly. Modest value claims are likely to be disproportionately expensive to investigate, let alone litigate, and so many will be uneconomic to pursue. The effective abolition of public funding in this area exacerbates these inequalities.
The limited scope of the strict liability provisions of the 1987 Act is unlikely to assist victims in many of these cases. I consider some of its deficiencies in my New Law Journal article, Driverless Vehicles: a future perfect? (Part 2) 30 November 2018. These include:
· The imprecision and subjectivity of the definition of ‘defect’ , based as it is on the reasonable expectations of the public at large test is too nebulous a measure and subject to change. Accidents contributed or caused by semi-automated / transitional stages of automation will attract great public interest in the media, which risks lowering the public’s expectations in this regard;
· The exclusion of wear and tear;
· The continuing uncertainty as whether all types of software are covered;
· The ability of manufacturers to issue lengthy disculpatory warnings in their manuals and literature, in the full knowledge that it is likely to be ignored by some (if not many) consumers;
· The relative ease with which a manufacturer like Tesla will be able to deploy the s4(e) defence that applies to cutting edge technology;
· The government’s failure to regulate mandatory PL cover and minimum levels of liability
· The ten year long stop limitation.
Furthermore, in addition to the inadequacies of this legislation, an individual claimant who is unable to establish strict liability under the 1987 Act will be faced with a forbidding evidentiary burden in establishing what was the true cause of the accident. It is conceivable that one of many individual component manufacturers would be responsible for a systems failure. It is equally conceivable that the software was susceptible to hacking by a malicious third party or that the cause of accident was contributed to the driver / user’s negligence . These are all unknowns (i.e. litigation risk) that undermine legal certainty and the prospect of an accident victims recovering his or her proper compensatory entitlement.
With all due respect, the inequality in access to justice that pervades the UK’s civil justice system is not something that any responsible government should ignore. Nor should it overlook the fact that by licencing these evolving levels of automation for use on roads it has actual or at least constructive knowledge of the correlative risks associated with their deployment. The government has a moral obligation to ensure not only that the vehicles it licences are fit for their intended use but also that victims are fairly treated and fully compensated if and when something goes wrong. The present artificial dichotomy been advanced automated vehicles at SAE Ls4 & 5 and all other forms of automation lacks moral coherence or other justification.
It is perhaps appropriate to note the government’s longstanding failure to fully implement Article 3 of Directive 2009/103/EC which requires any compulsory motor insurance for any civil liability (not just the personal fault of the user / owner / insured) and which extends to a much wider class of vehicle as well as private premises.
Urgent steps need to be taken to safeguard the interests of accident victims (which need to be given at least the same attention as the insurance industry). The most obvious solutions would be either to reform the CPA 1987 or to bring driver assist and conditional automation into the scope of the strict liability provisions of the AEVA 2018.