The Future of AI Liability in Europe

Image by VectorStock

Artificial Intelligence (AI) could revolutionise the world-wide economy as well as the way we live, work and interact with each other. While this new technology certainly presents great potential, it also comes with important risks to human life, health and wellbeing – among other risks.

In an effort to prepare for this new environment, the European Commission has been at the forefront of several initiatives that aim to provide a harmonised regulatory framework for the safe deployment and use of AI systems across Member States [1]. Amongst its most recent initiatives is a public consultation on how to adapt civil liability rules to the digital age and artificial intelligence. This public consultation, which closed on 10 January 2022, aimed to collect views on:

  1. how to improve the applicability of the Product Liability Directive (PLD) to the digital age, including AI, and
  2. whether there is a need to further harmonise rules of liability for damage caused by AI systems beyond the PLD.

The consultation is an important first step towards building a robust liability framework fit to address the current and future challenges posed by AI and the digital age in Europe. The changes that could be implemented as a result of the consultation could be immense and produce far-reaching consequences. Understandably, this public consultation attracted a high level of interest from various stakeholders, including businesses (Google, Bosh, Siemens, Avast), consumer organisations (BEUC, France Assos Santé), insurers (AXA, Insurance Europe, France Assureurs), NGOs, interest groups, legal scholars as well as members of the general public. In total, the European Commission received around 300 responses.

Pr. Jonas Knetsch (University of Paris 1 Panthéon-Sorbonne) and Dr. Emmanuelle Lemaire (University of Essex), assembled a small ad hoc research group, comprised of Pr. Michel Cannarsa (The Catholic University of Lyon), Dr. Laurie Friant (University of Paris 1 Panthéon-Sorbonne) and Pr. Simon Taylor (Paris Nanterre University), to produce a report in response to the consultation.

Overall, the authors of this report were of the view that the PLD should be adapted to enhance consumer protection in the digital age and increase legal certainty for all stakeholders. The authors also recognised that AI technology posed specific challenges and recommended that complementary measures be adopted to ensure the safe deployment and use of AI systems across Member States.

Adapting the PLD rules to the digital age and AI

The Product Liability Directive, which came into force on 30 July 1985, was a response to the increasing demand for consumer protection in a hyper-industrialised environment where goods were mass-produced, and mass-consumed. In essence, the Directive aimed to offer a high level of protection to consumers while ensuring that producers did not bear an undue burden. It was thus designed to strike a careful balance between the interests of both consumers and producers.

Yet, we must remember that the Directive was implemented at a time when the Internet was still in its early days, the use of AI remained largely theoretical, marketplaces were positioned in the ‘physical world’, and concepts such as ‘circular economy’ and ‘the Internet of Things’ (IoT) were simply non-existent. To say that the PLD – which did not undergo any major changes since 1985 – is in need of reform is certainly an understatement.

In order to adequately adapt the PLD to the digital age and AI, the authors of the aforementioned report took the view that the scope of application of the PLD should be extended, and in particular that:

  • the concept of ‘product’ should be expressly extended to intangible goods,
  • the concept of ‘producer’ should be extended to include online marketplaces and remanufacturers,
  • the concept of ‘damage’ should be extended to include specific types of immaterial loss (i.e. privacy or data protection infringements not already covered under the General Data Protection Regulation, and damage to, or the destruction of, data).

The authors of the report also recommended the amendment of specific PLD rules in certain situations, and more specifically:

  • the suppression of the development risk defence for AI products only,
  • the suppression of the 10-year longstop period in case of death or personal injury,
  • a clarification of the conditions enabling the 3-year limitation period to start running,
  • an alleviation of the burden of proof of ‘defect’ and ‘causation’ for products classified as ‘technically complex’ (which would include AI products and The Internet of Things).

In addition to recommending that the PLD be adapted, the authors of the report were also in favour of the European Commission adopting complementary measures in the context of AI to account for the specific features presented by this technology (autonomy, complexity, opacity, vulnerability, and openness).

Adopting complementary measures in the context of AI

The regulation of AI is proving challenging across legal systems, not least because of the difficulty in defining what AI is and what can be classified as an AI system. The European Commission made a recent effort to try and offer a clear – but open – definition of the term ‘AI system’ to ensure legal certainty while providing the necessary flexibility to accommodate any future technological developments. As the definition currently stands, an AI system means software that is developed with some specific listed techniques and approaches ‘and can, for a given set of human-defined objectives, generate outputs such as content, predictions, recommendations, or decisions influencing the environments they interact with.’[2] The definition is quite broad, and in consequence, the range of products based on – or using –  AI systems can be diverse and include voice assistants, image analysing software, search engines, speech and face recognition systems, as well as advanced robots, autonomous cars, drones or Internet of Things applications. Not all these products present the same type or level of risk, and some AI-based products are therefore more dangerous than others.

The authors of the report recommended that the European Commission consider:

  • the harmonisation of strict liability where AI-based products or services create a ‘serious risk of damage’ to consumers with an option to allow Member States to offer more protective liability rules to consumers,
  • the harmonisation of mandatory liability insurance for certain AI products,
  • the harmonisation of liability rules regarding the compensation of specific types of immaterial loss beyond the PLD (i.e. privacy or data protection infringements not already covered under the General Data Protection Regulation, and damage to, or the destruction of, data).

If you are interested in knowing more about the recommendations made by this university group to the European Commission, you can find a copy of their report (no. F2771740) – written in French – on the EC website or download it directly from our blog below:


[1] See e.g. European Commission, Communication from the Commission to the European Parliament, the European Council, the Council, the European Economic and Social Committee of the Regions – Artificial Intelligence for Europe (COM(2018) 237 final); European Commission, White Paper on Artificial Intelligence – A European approach to excellence and trust, (COM(2020) 65 final); European Commission, Communication Coordinated Plan on Artificial Intelligence (COM(2021) 205 final); European Commission, Proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on Artificial Intelligence (Artificial Intelligence Act) and amending certain Union legislative Acts (COM(2021) 206 final).

[2] European Commission, Proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on Artificial Intelligence (Artificial Intelligence Act) and amending certain Union legislative Acts (COM(2021) 206 final), Article 3(1).

Libel Trial against Investigative Journalist Concludes Before the High Court: A Landmark Test of the Public Interest Defence

Carole Cadwalladr speaks at TED2019: Bigger Than Us (April 15 – 19, 2019, Vancouver, BC, Canada) Photo: Marla Aufmuth via Flickr

By Alexandros Antoniou, Lecturer in Media Law, University of Essex

On 14 January 2022, a high-profile libel trial began before Mrs Justice Steyn at the Royal Courts of Justice in London. The British businessman Arron Banks sued investigative journalist Carole Cadwalladr for libel. Mr. Banks is an outspoken backer of Brexit. Ms Cadwalladr is an award-winning journalist who writes for the Guardian and Observer in the United Kingdom. She is particularly known for her work in uncovering the Cambridge Analytica scandal.

The case arose out of remarks in a Ted Technology Conference titled ‘Facebook’s role in Brexit – and the threat to democracy’ given by Ms Cadwalladr in April 2019, and a related Tweet. In the course of the Ted talk, which centred on the UK’s 2016 vote to leave the European Union, she said: “And I am not even going to go into the lies that Arron Banks has told about his covert relationship with the Russian Government”.

Arron Banks has always strongly denied any illegal Russian links, but he has admitted meeting Russian embassy officials on a number of occasions. Although his Leave.EU campaign was fined GBP 70,000 over multiple breaches of electoral law, the National Crime Agency’s investigation found no evidence of criminal activity.

Proceedings were issued on 12 July 2019. In a preliminary ruling on the meaning of Ms Cadwalladr’s words, Mr. Justice Saini held on 12 December 2019 that an average ordinary listener would have understood her words to mean: “On more than one occasion Mr. Banks told untruths about a secret relationship he had with the Russian Government in relation to acceptance of foreign funding of electoral campaigns in breach of the law on such funding.”

Mr. Banks maintained in his legal claim that the threshold of ‘serious harm’ under section 1 of the Defamation Act 2013 had been met in terms of damage to his reputation. Ms. Cadwalladr stated that this was not the meaning she had intended and that she had always taken care to say there was no evidence to suggest Banks had accepted any money. She originally pleaded the defence of ‘Truth’ under section 2 of the 2013 Act but, after Mr. Justice Saini handed out his ruling on the meaning her statement bore, Ms. Cadwalladr withdrew this defence in November 2020. She is now relying on the defence of ‘Publication on a matter of public interest’ under section 4 of the Act.

The defence under section 4 reflects principles established by previous case- law. It consists of two elements: Section 4(1)(a) requires that the words complained of were (or formed part of) a statement on a matter of public interest, and if the publication in question passes this test, then it also needs to meet the requirement of section 4(1)(b), which contains objective and subjective components.

The subjective component is that the defendant must believe the publication was in the public interest and the objective component is the question of whether it was reasonable for the defendant to hold that belief. Section 4(2) of the 2013 Act requires in particular that, in determining these matters, the court ‘must have regard to all the circumstances of the case’.

Thus, the central issue at this trial is likely to be whether it was reasonable for Ms. Cadwalladr to believe that the publication of her statements was in the public interest. The court will also look at the content and subject of the allegations, and the way the journalist acted in researching and reporting them. If Ms. Cadwalladr loses, she faces legal costs of up to GBP 1 million on top of damages.

In a piece published by Open Democracy, Ms. Cadwalladr stated: “Right now, we can’t police the money spent in our elections: this is a massive problem for our democracy. Facebook is unregulated and our electoral laws are still hopelessly unenforceable. There was (and still is) a huge public interest in journalists raising these issues – both as a warning for us here in Britain, and for countries everywhere”.

An interesting aspect of this case is that Arron Banks sued neither the Guardian Media Group which published Ms. Cadwalladr’s reporting for years nor TED which hosted her talk (or other large media outlets which made similar allegations). Instead, he chose to sue Cadwalladr personally. Press freedom groups have called for the case to be thrown out and described it as bearing many of the elements of a so-called SLAPP lawsuit – Strategic Litigation Against Public Participation. A key characteristic of such types of actions is the disparity of power between the claimant and the defendant.

The case has renewed calls for the UK Government to ensure that SLAPPs are not used to silence legitimate criticism and stifle any public interest reporting. Action to combat the emergence and growth of abusing litigation targeting journalists throughout the EU and ensure convergence in Member States’ approaches to SLAPPs is currently being considered at the EU level.

The Banks v Cadwalladr trial was heard over five days and judgment was reserved. The case has been followed closely by several investigative reporters. The Reporters Without Frontiers emphasised in particular that “the ruling will have serious implications for journalism not only in the UK, but internationally, given the popularity of London courts as a jurisdiction for such suits, and highlights the need for greater protections for journalists facing legal threats”.


This article was first published on the IRIS Merlin database of the European Audiovisual Observatory and is reproduced on the ELR Blog with permission and thanks.

The War on Compensation: Troubling Signs for Civilian Casualties in the Gaza Strip

Photo by Timon Studler

Dr. Haim Abraham, Lecturer in Law, University of Essex

Follow Dr. Abraham on Twitter here

The last round of belligerency between Hamas and Israel claimed a significant toll from civilians, with many arguing that some of the more devastating activities conducted by the IDF were in breach of the laws of war (for example, herehere, and here). Just days before a ceasefire was declared, Judge Shlomo Friedlander of Israel’s Be’er Sheva District Court released his ruling in the case of The Estate of Iman Elhamtz v. Israel, dismissing a claim for compensation for the killing of a 13 year old girl from the Gaza Strip by IDF forces in 2004. At first glance, this case seems to be just another instance in which the state’s immunity from tort liability for losses they inflict during combat is reaffirmed. However, a closer examination reveals that it is a significant development of the immunity, which could have vast ramifications for Palestinians’ ability to obtain compensation for losses they sustained from IDF activities that were in breach of the laws of war. Currently, Israel is immune from tort liability for losses it inflicts during battle, even if combatants inflicted the loss negligently. Yet, Judge Friedlander seems to expand the immunity further so that it applies not only to combatant activities that comply with the laws of war, but also to war crimes. This approach to the immunity has yet to be considered by the Supreme Court, but it is in stark opposition to international trends towards the scope of state’s immunity from tort liability.

The Elhamtz Case

The tragic events that resulted in Iman Elhamtz’ death unfolded in 2004 while Operation “Days of Penitence” was being conducted in the midst of the Second Intifada. An IDF military post at south Gaza Strip near Rafah was under high alert against threat of attack. Elhamtz approached the post, but the lookout did not properly execute his duties resulting in Elhamtz reaching within 100 meters of the post before being detected. Surprised and, according to their testimony, fearing that she is there as a part of a terrorist activity, soldiers began shooting towards Elhamtz even as she was fleeing. Elhamtz was killed. Major R, who was the commanding officer at the time, charged her dead body and engaged in ‘dead-checking’. A total of 20 bullets were found in her body.

A military court exonerated Major R of criminal wrongdoing. Elhamtz’ estate sought a ruling that would hold Israel accountable for her death through civil proceedings, and filed a tort law-suit against Israel in the Be’er Sheva District Court in 2005, arguing that she was shot and killed negligently and in violation of international humanitarian law.

In May 2021, the court dismissed their claim, despite finding that the military force indeed acted negligently and in violation of international humanitarian law. Judge Friedlander found that the military force was negligent on two counts. First, the lookout failed to perform his duties. If he was not preparing for a shift change but had properly observed the post’s surroundings, Judge Friedlander held, Elhamtz could have been spotted from a greater distance, chased away and probably would still be alive today. Second, the immediate and excessive use of force when there was no clear threat was in breach of the rules of engagement. The court adopted these rules to determine the relevant standard of care that is expected from a military force under such circumstances and held that this standard was breached. The military force should not have fired on Elhamtz to begin with, should have stopped when she began to flee, and dead-checking was completely incompatible with the standard of care that is expected from combatants.

The court also held that the actions of the military force violated the principle of proportionality. The sheer fact that Elhamtz was near the post, according to Judge Friedlander, does not mean that combatants can use deadly force against her. Even if she posed a risk, which was highly doubtful, she should have been chased away or restrained, not killed.

The sole reason for which Israel was not held liable for the death of Elhamtz was that Israel, like many other countries, has a special immunity from tort liability for losses it inflicts during armed conflict called the ‘combatant activities exception’. Through his opinion, Judge Friedlander paved the way to reject future tort claims that are likely to be filed by Palestinian casualties from the most recent round of fighting. But to understand the legal mechanism that allows this reality, a better appreciation of the immunity is needed.

The Combatant Activities Exception

In the mid-20th century, states began reforming laws concerning their immunity from tort liability, by removing procedural and substantive hurdles for filing claims, as well as limiting the scope of the doctrine of sovereign immunity to enable holding foreign states liable in tort. Nevertheless, while immunity from liability became more limited, it was not done away with altogether. Some pockets of immunity remained, including the combatant activities exception, which, essentially, provides a blanket immunity from tort liability for wrongful actions conducted in battle.

The scope of the combatant activities exception varies between jurisdictions. Canada, for example, has what appears to be the broadest statutory exception, which precludes liability for “anything done or omitted in the exercise of any power or authority exercisable by the Crown, whether in time of peace or of war, for the purpose of the defence of Canada or of training, or maintaining the efficiency of, the Canadian Forces.” The U.S. statutory exception is somewhat more limited in its scope, maintaining that no liability would be imposed in “any claim arising out of the combatant activities of the military or naval forces, or the Coast Guard, during time of war.”

When Israel first enacted its version of the combatant activities exception through the Civil Wrongs (State Liability) Act 1952, it was very similar to the U.S. exception, simply stating that Israel is not “liable in tort for a combatant activity committed by the Israel Defense Forces.” Initially, courts interpreted the exception narrowly, holding that it is applicable only to activities in which there was an objective and immediate risk that is of a combatant character. However, with each major conflict with the Palestinian population, the scope of the exception was expanded through judicial interpretation and by legislative amendments. These expansions have three notable themes.

First, the boundary between combatant and non-combatant activities has been blurred. During the First Intifada (1987-1993), IDF forces faced large-scale violent protests. Policing operation within the Occupied Palestinian Territories exposed the forces to imminent risk to their lives, and courts were torn between a narrow and a broad interpretation of the combatant activities exception. The narrow approach ruled out the exception’s applicability, holding that policing activities are not combatant activities, even if they are conducted by military forces who are exposed to considerable risks. The broad approach held the contrary view, finding that the exception is applicable even for policing activities due to the real risk to soldiers’ lives, who were operating in a hostile environment. Ultimately, the broad interpretation of the combatant activities exception was adopted by the courts and the legislature, expanding the scope of the exception to include policing and counter-terrorist activities. The exception became so broad that it currently applies to activities in which a soldier subjectively (and mistakenly) feels at risk, as were the circumstances that led Judge Friedlander to hold that Israel cannot be held liable for the killing of Iman Elhamtz.

Second, non-Israeli Palestinians are viewed as ‘the enemy’, and their tort claims are thought of as a continuation of terrorist activities through civilian means. For example, in 2005 the Israeli parliament sought to expand the scope of the exception to include any and all injuries in the Occupied Palestinian Territories, whether combatant or not. This legislation was struck down by the Supreme Court on the grounds of it being unconstitutional. Similarly, a 2012 amendment expanded the applicability of the exception to all non-Israeli residents of the Occupied Palestinian Territories, regardless of the character of the activities that resulted in the loss (this legislation is currently being reviewed by the Supreme Court).

Third, the temporal and geographical distinction between war and peace has been undermined. The original interpretation and definition of the exception meant that it was applicable only to engagement in actual hostilities. The courts examined the circumstances that resulted in the injury, analyzing the particular location in which the activity took place, taking into account a small window of time before or after it. Following the events of the Second Intifada (2000-2005), as well as the legislative expansion of the exception, courts consider an ever-increasing context. Courts no longer examine just what happened on the particular street and time in which someone was injured. Instead, they refer to the general area and history to deduce whether combatants faced a risk that would fall under the scope of the exception, sometimes expanding the timeframe to years prior to the activity that resulted in the injury itself.

The overall effect of the expansion of the combatant activities exception resulted in a dramatic decrease in the number of tort suits being filed, from thousands of cases in the early 2000s to a handful of cases a year currently, and a finding of liability against Israel is nearly impossible. Nevertheless, the scope of the exception is still being contested by plaintiffs, and it is far from clear that its current form can be Justified (see, for example: hereherehere, and here).

‘Testing the Waters’

The dismissal of the Elhamtz case coincided with the growing criticisms of Israel’s violations of the laws of war during the 2021 round of belligerency between Hamas and Israel. These should have been two unrelated matters. One revolved around a tragic incident in 2004, the other was still ongoing in May 2021. Yet, Judge Friedlander’s opinion, which held that the exception applies not only for the military’s negligent actions, but also for its actions that violate international humanitarian law, seems to create a link between the two. In the obiter, Judge Friedlander gave contrasting examples to illustrate the limits of the combatant activities exception, noting that even if one country indiscriminately and disproportionately bombs the civilian population of another country during an armed conflict, it is a combatant activity for which the exception applies.

Judge Friedlander did not need to use this example to reach the conclusion that the exception applies. The Supreme Court has ruled years ago that claims for compensation for violation of international law should be pursued through separate proceedings, not through tort claims, and that the exception applies even for negligent injuries by the IDF. Invoking this particular example at that particular time does not appear to be a redundant hypothetical, but rather laying the groundwork for dismissing future claims that are bound to be filed against Israel for the losses it inflicted in 2021.

The Supreme Court has yet to give clear guidance on whether the combatant activities exception can apply when the State’s actions are in clear violation of the laws of war. There is a growing trend in the international community to limit the availability of states’ immunities in such cases. If the Supreme Court of Israel was to adopt Judge Friedlander’s approach, it will be expanding the scope of the combatant activities exception significantly, blurring the line between legitimate combatant activities and criminal activities. Such an interpretation appears to contradict the position that was raised in several obiters by Israeli courts. On various occasions, courts clarified the limits of the combatant activities exception by stating that criminal activities, such as looting, do not fall under the combatant activities exception even when they are done on an active battlefield. It is hard to find a rationale that will allow for an imposition of tort liability for looting property but not for committing war crimes. Neither is a legitimate act of war, and both should be excluded of the dispensations that accompany sanctioned warfare.

This post first appeared on the Blog of the European Journal of International Law and is reproduced on our research blog with permission and thanks. The original article can be accessed here.

Serial Health Risks and Civil Liability: A Comparative Study of French and English Law

Photo credit: Pixabay

Dr. Emmanuelle Lemaire, Lecturer in Law, University of Essex

Abstract

Serial health risks are increasing in occurrence and have the capacity to affect large numbers of people living in different countries. Thalidomide, contaminated blood, Diethylstilboestrol (DES), asbestos and Bovine Spongiform Encephalitis (BSE) are among some of the most well-known examples of serial health risks damaging human lives, but the recent Covid-19 outbreak reminds us that such risks have always existed, primarily in the form of epidemics. Yet, the legal treatment, in civil liability, of personal injury resulting from serial health risks often lacks clarity not only because serial health risks have diverse origins, but also because they can trigger different regimes of civil liability. Given their increasing importance and this lack of clarity, I devoted my PhD thesis to the comparison of how the French and English legal systems responded to serial health risks in civil liability. I concluded that while these two legal systems are facing similar legal challenges from a civil liability perspective, they have not always responded in a similar manner. In fact, I argued that their differences of approach highlight unreconcilable views as to the conception of civil liability, its role and its underlying objective. My analysis also led me to conclude that despite these differences, the legal treatment of victims of serial health risks remained unsatisfactory in both legal systems due to unjustifiable inequality of treatment. The following post will provide a summary of the key-findings from my PhD thesis.

Background

My interest in comparing the legal treatment of serial health risks in French and English civil liability stemmed from a theme that I uncovered while I was writing my master’s dissertation in France. At the time, I was working on issues of proof of causation in civil environmental liability in French and English law and discovered that, besides environmental cases, other cases raised similar issues of proof of causation.

Yet, these cases involved risks with diverse origins: the risks were occupational (asbestos, brick dust, mechanical vibrations, etc.), technological (electromagnetic waves, nuclear radiations), pharmaceutical (thalidomide, contaminated blood, DES, PIP prostheses, metal-on-metal hip prostheses, anti-hepatitis B vaccine, pertussis vaccine, etc.), food-related (BSE-contaminated beef), consumer-related (tobacco), industrial (Round-up) and environmental (pollution). More importantly, these cases triggered the application of different civil liability regimes: liability for defective products, nuisance, and negligence (employers/employees, producer/consumer, etc.).

Nevertheless, despite these important material differences, they also presented common features: claimants always suffered from diseases, a specific type of personal injury, and as a result, all cases were concerned with risks to human health; they were all connected to the widespread use of products and substances, therefore raising the number of victims to potentially catastrophic levels; finally, all involved serial damage, a legal concept sometimes used by continental legal scholars to refer to a type of damage that affects a number of people spread in various locations and at different times but connected to a common origin.

In summary, these cases were all dealing with health risks leading to serial damage which I will refer to as ‘serial health risks’. Given their common features, it was no surprise to observe that they often raise similar issues for the application of civil liability.

As I already mentioned, proof of causation is often a crucial issue in these litigations because our understanding of aetiology of diseases is generally fraught with significant gaps in scientific knowledge. As a consequence, the existence of scientific uncertainties is particularly challenging to the success of tort claims. Additionally, when numerous plaintiffs are similarly affected by the same substance or product, courts are likely to face a flood of tort claims. Large cohorts of claimants put a strain on the justice system, thereby leading to question whether the tort system is the best way forward in dealing with compensation, and if so, how it copes with large numbers of claimants.

Research question

While this area of litigation has become increasingly important both in the number of cases and in monetary value (potentially multi-millions pounds litigation), legal scholars have mostly purported to follow a sectoral approach, devoting their analysis to one legal issue at a time (causation, collective redress, etc.). I believed that a more comprehensive study of the interactions between serial health risks and civil liability was needed in order to bring clarity into this complex area of law and that drawing on the legal experiences of both the French and English legal systems could be beneficial to inform on how similar issues of civil liability were approached in legal systems with different legal traditions.

Through this comprehensive analysis, I sought to investigate how personal injury resulting from serial health risks was treated by civil liability in both legal systems.

Choice of jurisdictions and methodology

The French and English legal systems have been chosen because they have different legal traditions (civil law vs common law) and were influenced by European Union law. More importantly both legal systems were good comparators because they have had to deal with the same issues of serial health risks at a similar point in time using the same scientific knowledge available, examples being contaminated blood, asbestos or BSE. For this reason, the comparison of French and English responses to this new problem seemed particularly apposite.

To conduct my research, I borrowed from both functional and contextualist methods of comparison. When looking at the practical issues connected to the application of civil liability in the context of serial health risks and the solutions developed in French and English law to address them, I used a functional method through a careful examination of French and English jurisprudence. However, I believed this approach to be limited: while the solutions adopted in both legal systems could be functionally equivalent, they could nonetheless be justified by different reasons. Hence the real divergences between legal systems could well remain hidden if my analysis was limited to a functional approach. I therefore complemented my comparative work with an analysis of political and social contexts to identify the underlying reasons and objectives behind the adoption of solutions in French and English law.

Comparative findings

My main findings can be summarised as follows:

  • The comparative analysis revealed that French and English law are moving towards improving the protection of victims’ interests in the context of serial health risks.

For example, both legal systems introduced compensation schemes to provide victims of certain serial health risks with expedient compensation under more flexible requirements than that of civil liability. Such schemes have benefited victims of contaminated blood, asbestos, vaccines (France and England), Diethylstilboestrol (DES), Mediator, Depakine (France), and BSE (England). Furthermore, French and English legislators have introduced strict liability regimes in sectors of activity where serial health risks could materialise with the aim to facilitate defendants being held legally liable. The introduction of the nuclear liability regime and the liability for defective products are the most seminal examples (chapter 3).

Similarly, I show that collective redress mechanisms have been introduced in the French and English legal systems not only to cope with large cohorts of claimants seeking compensation through judicial process but also as a means to restore the ‘balance of power’ between plaintiffs (private individuals) and defendants (often large companies) through allowing individual plaintiffs to aggregate (Goliath v. Goliath). I demonstrate that the conditions to launch group litigations are designed to deal with victims of mass torts which include victims of serial health risks, although I also identified difficulties in both legal systems which impede the efficacy of these mechanisms in practice.

Finally, the analysis of case-law provided me with clear evidence that proof of the traditional requirements of civil liability (in particular causation) has sometimes been adapted to allow victims of serial health risks to succeed where they should have normally failed (chapter 5 to 10). Once again, this is indicative of the importance given to victims’ interests from a civil liability perspective.

  • However, despite these converging evolutions, I conclude that the treatment of serial health risks in French and English civil liability remains fundamentally different.

For example, the comparison of French and English compensation schemes in the context of serial health risks reveals that while French schemes are often designed as an alternative compensation mechanism to the tort system, English statutory schemes are engineered as a supplementary compensation mechanism to the tort system. In other words, where French compensation schemes generally compete with the tort system, English statutory schemes are created as a last resort mechanism and can only be actioned when all else fails. This important divergence underlines the fact that the French legal system readily separates compensation from civil liability in order to promote compensation while the English system continues to see the tort system as the primary route for obtaining redress. Similarly, while legislators on both sides have introduced strict liability regimes, their number remains limited in England while in France they continue to grow. This suggests a willingness by the French legal system to move from fault-based liability to strict liability. In contrast the English legal system is reluctant to depart from the fault principle to compensate personal injury resulting from serial health risks (chapter 3).

The application of tort law rules to situations of serial health risks is another striking example of the different approaches taken by the two legal systems. In chapter 2, I developed a detailed typology of serial health risks based on the level of scientific knowledge and distinguished ‘known’, ‘suspected’ and ‘unknown’ risks. In chapter 5 to 10, I used this typology to examine the judicial treatment of serial health risks in French and English civil liability. In French law, I found that the conditions of civil liability are consistently adapted, and that proof of these requirements are facilitated to ensure that the victims could obtain compensation, regardless of whether the situation is one where the risk is unknown, suspected or known. In other words, French law favours victims’ compensation in all situations of serial health risks. Conversely, the position is much more nuanced in English law: the less knowledge that is available as to the existence of risk, the less likely it is that relaxation of tort law will be accepted (chapter 6 to 10). As a result, proof of causation has been adapted where risks are known but not where risks are either suspected or unknown.

  • I argued that the aforementioned differences are not simply differences of degree between the legal systems. Rather they show irreconcilable views as to the role played by civil liability in the context of serial health risks.

When considering the balance between the protection of bodily integrity (claimants’ interest) and the protection of economic freedom (defendants’ interest), the French legal system took the clear view that the protection of bodily integrity is to be favoured in the context of serial health risks. To this effect, civil liability is simultaneously declining due to other compensation mechanisms emerging, and yet strengthened due to the introduction of strict liability regimes, collective redress mechanisms and relaxation of tort law conditions. These conflicting trends can all be explained when regarding the underlying goal being pursued in French law, which is ensuring that victims of serial health risks obtain compensation. The interests of defendants are therefore systematically sacrificed.

The English legal system has made another choice. While the protection of bodily integrity is undeniably important, it is not a sufficient reason, in and of itself, to sacrifice the defendants’ interests. Other reasons are needed to support a relaxation of the traditional rules of civil liability and examination of English case law reveals that judges are careful to evaluate conflicting interests to determine which one should be given precedence. This explains why the more scientific uncertainty there is with regards to the existence of the risk, the less willing English law is to relax the traditional rules of tort law. This approach reveals that English law is much more concerned than French law with the need to strike an appropriate balance between the protection of bodily integrity and the protection of economic freedom.

  • Finally, I found the legal treatment of serial health risks to be unsatisfactory in both France and England, albeit for different reasons.

If, as my analysis suggests, the French legal system is overly concerned with the compensation of victims of serial health risks, then the arrangements currently in place are not satisfactory, especially when considering the solutions crafted through legislature. French legislators have succumbed to the temptation to develop solutions on a sector-by-sector basis which can lead to victims being treated differently simply because serial health risks have different origins. For example, French legislators introduced collective redress for victims of personal injury in the health sector but not in relation to the food industry. As a result, thalidomide-type victims could benefit from group litigations but not BSE-type victims. I found this sectoral approach to be sorely lacking and difficult to justify.

The English legal system has favoured an approach whereby victims of serial health risks are treated differently depending on the level of scientific knowledge available. I argued that this approach can be justified and is more rational than that adopted in French law which differentiates victims according to the origin of serial health risks. Yet, the English approach is not without problems because it can also lead to victims of serial health risks being treated differently for no valid reason. English judges are trying to preserve a balance between conflicting interests (bodily integrity vs economic freedom) which results in relaxation of traditional rules to have limited applicability. In order to ensure limited applicability, English judges are drawing subtle distinctions between cases. For instance, the applicability of exceptional rules of causation (chapter 9 and 10) depend on whether the situation is one of alternative or cumulative causation, whether there is one or multiple noxious agents at play, whether the disease is divisible or indivisible. Hard cases often challenge the practicality of these distinctions and reveal that, as is the case in France, victims of serial health risks can be treated differently with little rationality behind the solutions. The worst example is linked to victims of asbestos where the solutions depend on whether such victims are suffering from mesothelioma, a specific disease linked to asbestos exposure, or not.

Conclusive thoughts

This comparative analysis was designed as a starting point. It aimed to provide a clear picture of the current position of how French and English law are dealing with the fallout of serial health risks from a civil liability perspective. Shortcomings have been identified in both legal systems which need to be addressed in future work to ensure that victims of serial health risks are treated more equally. The recent Covid-19 outbreak, probably the biggest serial health risk that our generation had to face, serves to reinforce this point.

Essential workers have put their lives on the line to allow for essential activities to continue during the lockdown. Healthcare professionals and carers primarily, but also supermarket workers, train and bus drivers, firefighters, postmen and delivery men amongst many others have had to work, sometimes in very close proximity to the general public, in order to deliver these essential services. Some of them, as we know, have died due to Covid-19 therefore raising the question of whether the State should step in to provide some type of compensation to essential workers and their bereaved families. It would appear that, once again, not all key workers have been treated the same. In particular, healthcare professionals and carers have benefited from special arrangements both in France and in England, a “privilege” that was not extended to other key workers.

While the nature of the work conducted by healthcare professionals and carers undeniably put them at a high risk of contamination from Covid-19, the same could be said of all key workers who have had to work in contact with the public. Why then should some essential workers be treated better than others? This latest example shows that unless we are planning on rethinking our general approach to serial health risks, the shortcomings previously identified in my work will be further aggravated. 

This post originally appeared on the British Association of Comparative Law Blog and is reproduced here with permission and thanks.