​

​

Civil Liability for the Use of Electronic Forms and Mechanisms of AI,

inter alia in the Sphere Of Transport

​

on the basis of the European Legal Framework

The innovations powered by Artificial Intelligence are deployed in miscellaneous aspects into modernity. The scientific article focuses on the necessity of the implementation of the ad-hoc regulation on AI. The granular approach for the note of Artificial Intelligence and the specificity of providing ‘electronic personhood’ are reviewed. The tendencies in the legislations’ enhancement and alterations are enlightened, with a proposed corresponding solution to particular issues that may arise.  The international taxonomy in the civil liability for the use of automated vehicles is analyzed, and the reasons for illuminating the relevant subject are substantiated.

​

Keywords: artificial intelligence, civil liability, automated vehicles, innovations, transport, electronic personhood, legislation on AI.

​

Formulation of the relevance of the scientific article. Technological advancement has led the humankind to the threshold of an aeon where various sophisticated electronic forms and mechanisms with Artificial Intelligence (henceforth referred to as ‘AI’) implemented are becoming applicable. Such innovations imply comprehensive alterations, pertaining to the functioning of the public and social relations which, consequently, are being affected by technological novelties. European Parliament in its ‘Resolution of 16 February 2017 with recommendations to the Commission on Civil Law Rules on Robotics’ formulated substantive reasons for the legislature to take into account the technological implementation’s implications of a different nature and effects on diverse areas of life with an aim of guaranteeing respect towards people’s physical wellbeing, health, safety and security of human rights, ensuring transparency and preserving the economy’s stable functioning [1]. 

​

In particular, significant dimensions of the motives such as below-stated were mentioned:

• 1) the necessity of creating a complex definition for robots and AI in order not to hinder the innovative progress;

• 2) the regulation of intelligent machines in the long-term which may include independent decision-making functions;

• 3) the assessment of the impact of technologies on employment where there is a probability of increase of the unemployment rate in case of the improper regulation of AI’s advancement processes which may lead to abrupt transformation of the labour market into a human-disoriented cluster, humans being vulnerable to automation;

• 4) the importance of the understanding that technological development is likely to curtail the percentage of the middle class and, as a result, its influence, concentrating the power under control of a minority;

• 5) bearing in mind the necessity of establishment of legal liability of those who are involved in the elaboration of AI manifestations for the quality of such products.  

 

Furthermore, concurring the views of the European Parliament, it is also vital to give prominence to determination of the liability not only for the manufacturers, but the operators, the owners and the users of AI forms and mechanisms with the purpose of guaranteeing the fundamental human rights, laid down in the Universal Declaration of Human Rights, proclaimed by the United Nations General Assembly and adopted by Ukraine in particular [2].

​

Recent research and publication analysis. Certain provisions that reflect the changes AI has brought into life are quite novel. In our research we are primarily focused on the European legal framework which is undergoing the stage of formation parallel with the scientific works of respected scholars. We have taken into consideration the process of formulating the definition for AI, the necessity and consequences of the absence of the respective regulation thereof were enlightened by J. McCarthy, M. L. Minsky, N. Rochester, C.E. Shannon, Stephen Hawking, also studying the documents issued by the European Parliament as well as the European Commission. Specifically, we made allowance for the significant need for the regulation of AI implemented into the realm of the advanced transport system, especially in the perspective of civil law, studied by the EU institutions and the following scholars: Karolina Ziemianin, Sahin Ardiyok and Armanç Canbeyli. Moreover, the Ukrainian Concept on the development of AI and the new Civil Code provisions envisaged in the relevant Project were taken into account.

​

In consideration of the level of the technological development in the realm of robots and AI, the main purpose of this scientific article lies in giving prominence to the civil liability aspects, specifically in the sphere of transport with AI implemented.

​

Presentation of the main body of the article. In the first instance, we are going to elaborate on the notion of Artificial Intelligence as, how it is characterized in the study, commissioned by the Policy Department C at the request of the Committee on Legal Affairs, ‘AI is a new buzz word’. Hence, a great variety of definitions has been given to AI devices. Here, we might put the emphasis on the twofold purpose of the paramount importance of defining what AI really is: firstly, determining AI stricto sensu, understanding its nature in the matrix of technologies and secondly, figuring out whether the policy makers and legislatures are given the commensurate and relevant interpretations of AI. 

​

As the criteria of defining AI are uncertain and inadequate, this could lead to infliction of harm on humankind and ineffective, no longer in tune with realities regulations. Moreover, such cutting-edge novelties may possibly be classified into various types of devices with specific rules for each of them as they are intrinsically bound with diverse technologies in a variety of ways. For instance, a driverless car, a smart-toothbrush, a robot-companion and a non-embedded expert system for medical diagnosis are not commensurate with one another in full measure [3].

​

While agreeing with the aforementioned study, that ‘only by abiding a granular approach it is possible to achieve the level of accuracy, necessary to understand the technology, the issues it might give rise to, and ultimately the domains that require regulation’, we are going to focus on the fundamental concept of the AI’s introduction. The term ‘Artificial Intelligence’ was proposed by J. McCarthy in 1956 during the Dartmouth Artificial Intelligence conference which is often considered as the watershed moment not only in computer science, but also in psychology, mathematics, engineering and many other realms. The purpose of the research project was crystallized as ‘to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it’ [4]. 

​

50 years later, John McCarthy who was working in Stanford University, stated in his article that AI ‘is the science and engineering of making intelligent machines, especially intelligent computer programs’ [5]. Such practice points to the conclusion that AI is taken into consideration as science or the branch of computer science, but it can also be claimed that such approach of determining AI is quite debatable for the purpose of defining the system in the legal framework. 

​

Analyzing the study by Policy Department for Citizens’ Rights and Constitutional Affairs which has already been mentioned, taking into consideration different policy making definitions of AI provided, we may give the acknowledgement that AI is a sophisticated phenomenon which yet has no general definition applicable all over the world. What is more, the willingness to formulate an all-encompassing definition for AI is clear, but is almost impossible on an empirical basis as nowadays the technologies are being ameliorated, thus, it is hard to foresee the future of innovations, their forms and impact on the welfare of mankind. For this reason, even after accepting the generic notion of AI, it may be complete and fully effective for a short period of time. That is why a general definition is advisable to formulate a cornerstone of AI (for instance, to treat AI as ‘an organized set of information technologies’ with due regard for the Concept on the development of Artificial Intelligence in Ukraine) [6], but multiple sides of the innovation demand the according ‘sector-specific approach that does not prioritize the technology, but focuses on its application within a given domain’ [7]. In a regulatory perspective, it will provide a certain taxonomy based on the specific features of the AI implemented, technology’s use and its implications on social and economic relations, will concretize specific aspects of AI and solutions to the relevant issues that may arise, and thereby make the legislation more adaptable to the operational amendments.

​

Being cognizant of the legislators’ ubiquitous inertia, there is a need in obliging the countries to provide alterations to the legal framework which shall be enacted simultaneously or immediately following the introduction of new technologies for the purpose of ensuring the balance betwixt the innovations and an adequate liability as well as the overall harmonization of the legislation.

​

Our scientific article would not be comprehensive without mentioning the proposal given by the European Parliament in the 2017 ‘European Parliament Resolution on Civil Law Rules on Robotics’ which envisages a creation of a specific legal status for robots in the long run. Without a doubt, the bold statement has given the rise to heated debates, where on the one hand, robots may interact with third parties independently, making their own decisions; while on the other hand, it can be claimed that human supervision and control take action and directly influence the development of the intelligence of the robot, thus, a human being has to be liable for the robot. 

​

Later, in 2019, the Expert Group on Liability and New Technologies appointed by the European Commission criticized the ‘electronic personhood’, inter alia because ‘there is currently no need to give a legal personality to emerging digital technologies’ and ‘any sort of legal personality for emerging digital technologies may raise a number of ethical issues’ etc. 

​

On this matter, we take into consideration a similar position enlightened in the holistic research article ‘Civil legal personality of artificial intelligence. Future or utopia?’ written by Karolina Ziemianin, where the conclusion has been made that ‘there are currently no grounds to grant legal personality to artificial intelligence. Therefore, liability for damages caused by artificial intelligence must be borne by natural persons, legal persons or organizational units with legal capacity. The principles of this responsibility should depend on the type of artificial intelligence and its technological advancement’ [8]. 

​

However, the rapid progress of trailblazing projects in the technological sphere may prompt to the aforementioned idea of conferring legal personality to AI innovations; but even if self-sufficiency of AI can be assumed, the people and technologies shall be jointly and severally liable as required by law. In case of a defective program, confirmed by specialists, the technology must be disbanded or eliminated. It is vital note, by the ‘joint liability’ and direct destruction of certain technologies we accentuate not on the subject of responsibility that can be held by the digital device, but the foreseeable burden of those parties who possess rights over them. It also has to be emphasized that the term ‘electronic personhood’ doesn’t equate human beings with technologies that stand as product of human intellect and shall not take the forefront, gaining a greater priority than a person.

​

Thus, if Artificial Intelligence was more independent and capable than humans, having an opportunity to pursue its own intentions, such intrinsically perilous devices would have to be banned. Professor Stephen Hawking, prominent British scientist, properly mentioned in one of the BBC interviews that ‘The development of full artificial intelligence could spell the end of the human race’ [9]. As we consider the innovations that can be taken under legal control and the responsibility of people therefor (precisely in a civil law vista – it is a burden with compensating damages), the differentiation of liability provisions has to be made, in view of some functional rationales which would reflect the fundamental guarantees of human rights and freedoms and the overseen technological advancement. Hence, we do not exclude the option of ‘electronic personhood’, the issue thereof may become a relevant legal dilemma in the nearest future which already has to be examined and researched.

​

As the European Union highly values the development and improvement of the urban environment’s functioning, one of the great varieties of initiatives that is intrinsically intertwined with the subject of our article is the idea of so-called ‘smart cities’ that prescribes the eventuality of upgrading the respective networks of transport. The targets of upgraded cities lay in amelioration of the citizens’ life, focusing on their main needs and reasonable preferences, one of which include the automated road mobility, supposed to contribute to safety on the roads among other things.

In such a state of affairs, the perspective of making Europe a world leader in the deployment of connected and automated mobility seems to be quite logical and incremental. This ambition was proclaimed by the European Commission in a strategy paper ‘On the road to automated mobility: An EU strategy for mobility of the future’, one of the aims of which lies in reducing the cases of fatalities on the road [10].  Hence, the European fully autonomous car market is evolving rapidly and more and more vehicles aforementioned are supposed to appear on highways in the nearest future [11].

​

We find it important to note the general definition of an automated vehicle provided by the European Parliamentary Research Service, where it is identified as ‘a motor vehicle (car, truck or bus) which has technology available to assist the driver so that elements of the driving task can be transferred to a computer system’ [12]. In comparison, an autonomous vehicle doesn’t require any human intervention and is classified under the 5th level of driving automation of SAE [13]. 

​

In view of the facts set above, the legal aspect concerning the establishment of a civil liability for the damages, caused by the use of automated vehicles, is a matter of crucial importance that we are enlightening in this scientific article. It has to be emphasized that in the system of the civil liability two principal goals have to be achieved – compensation and deterrence of the unfavorable behavior. In order to comply with such aims, potential AI-fault parties have to be considered.

​

Having reviewed the established practice, in car accidents that involve human drivers, the party which held the greatest responsibility for the accident is usually liable to compensate the other party (viewing the application of a fault-based framework in the form of negligence). However, for vehicles with a high level of automation, where the level of human intervention is minimized, the definition of the civil liability is gaining in its complexity. The generally accepted notion of a human driver starts lacking clarity in establishing future provisions of liability.

​

It is significant to highlight that in the Project of the Update of Concept of the Civil Code of Ukraine there are statements, specifying the re-examination of the system of special delicts, inter alia reparation of damages caused by Artificial Intelligence [14].  In this case, we shall specify manifold potential at-fault parties: the driver or owner of the automated vehicle; the automated technology manufacturer; the dealership that sold the automated vehicle; component and materials manufacturer of automated vehicle etc.

​

In the previous research ‘Problematics of Legality in terms of the AI’s vehicles global deployment’ it was accentuated that in accordance with Vienna Convention on Road traffic, article 8 (5) sets forth that: «every driver shall at all times be able to control his vehicle» [15]. Thus, if it is proven that the car accident has been caused by the driver’s negligence, the person, who has been behind the wheel, has to be liable. In most researches we studied, there have been no differentiation between the automated vehicle’s owner and exactly the user of the vehicle common in the EU. The reason for such practice is that ‘at the EU level, mandatory insurance requirements are set forth by the Directive 2009/103/EC on the insurance against civil liability in respect of the use of motor vehicles, which imposes compulsory third-party liability insurance on the owner for damages to property, along with loss and personal injury inflicted on another party because of the actions of the policyholder. Such third-party insurance covers damages caused to pedestrians and all passengers, other than the driver (art. 12 MID)’ [3]. 

​

Nevertheless, the particular automated vehicle may be used not only by its owner, but by other persons – from our point of view, this circumstance has to be taken into account and hence, the liability of the driver and the possessor of an automated vehicle shall be delineated in order to protect the rights of the latter person. What is more, such a legal model that implies preventative measures by attaching the liability to the driver, not the owner of the automated vehicle may compel namely the user of the vehicle to diligently comply with the road traffic regulations. Our position can be supported with certain provisions enacted by Germany in 2017 where it is prescribed that the demeanor of the driver of a high or fully automated vehicle up to the level 4 SAE is regulated, giving the latter the duty to resume control over the vehicle every step of the way when the system identifies such a need or particular objective circumstances arise. The case of non-compliance with the rule aforementioned would entail the driver’s fault-based liability. The owner, in such a legal situation, shall be liable in the event that the driver’s fault is absent.

​

The company which brings the idea of automated vehicles into life may be liable if the damages are caused by a defective product. However, when the producer cannot be identified, the dealership of the defected automated vehicle is supposed to take responsibility. It is also vital to focus on the complexity of the vehicle’s development, where some experts highlight two main elements of the failure’s causes – a defect in software or the product itself.  Thus, a few parties fully engaged into installing the AI software into technological innovations (e.g.  the software companies) may contribute to the product manufacturing as well and be liable thereafter.

​

It is significant to acknowledge the prognosis which comprises the possibility of automated vehicles’ hacking in the nearest future. The public danger of the car theft may exponentially increase in liaison with an absolute unpredictability of the actions of the autopilot vehicle powered by AI. The exclusions or restrictions of the liability in case of an unauthorized use of the vehicle are traditionally prescribed for ordinary vehicles (provided that the owner had not enabled such a situation happen by his or her own careless conduct). However, in terms of the AI’s deployment, the sensitization on this subject has not been given the relevant attention yet. Thus, it shall be clarified and stipulated in the legislation, drawing attention of the programs’ developers to ensure a high degree of the automated vehicle’s protection and giving no prospect for perpetrators to seize the opportunity of the legal lacuna.

​

Taking the research material into consideration, a conclusion can be made that the modern world is endeavoring for advancement by leaps and bounds and the legislation has to keep up with such tendencies. Hence, the adoption and the relevant harmonization of the ad-hoc AI legislation at the EU level, together and simultaneously with other countries all around the world is currently needed. Furthermore, the clarity of the parties’ liability would facilitate the access to justice of potential victims and propel success of the automated vehicles at the market, providing the protection of fundamental human rights, reliance, transparency and permanence during the times of tectonic and crucial changes we are becoming the witnesses of.

​

References

​

1. European Parliament resolution of 16 February 2017 with recommendations to the Commission on Civil Law Rules on Robotics (2015/2103(INL)) – URL: https://www.europarl.europa.eu/doceo/document/TA-8-2017-0051_EN.html, Accessed on 24 November 2021

​

2. Universal Declaration of Human Rights – URL: https://www.un.org/en/about-us/universal-declaration-of-human-rights, Accessed on 24 November 2021

​

3. ‘Artificial Intelligence and Civil Liability, Legal Affairs’, study requested by the JURI committee – URL: https://www.europarl.europa.eu/RegData/etudes/STUD/2020/621926/IPOL_STU(2020)621926_EN.pdf, Accessed on 24 November 2021

​

4. A proposal for the Dartmouth summer research project on Artificial Intelligence – URL:  http://www-formal.stanford.edu/jmc/history/dartmouth/dartmouth.html, Accessed on 24 November 2021

​

5. ‘What is Artificial Intelligence?’, John McCarthy – URL:  http://www-formal.stanford.edu/jmc/whatisai.pdf, Accessed on 24 November 2021

​

6. Concept on the development of Artificial Intelligence in Ukraine – URL:  https://prianykovabusiness.wixsite.com/defender/projecto-1, Accessed on 24 November 2021

​

7. AI Now Report 2018 – URL: https://ainowinstitute.org/AI_Now_2018_Report.pdf, Accessed on 24 November 2021

​

8. ‘Civil legal personality of artificial intelligence. Future or utopia?’, Karolina Ziemianin – URL:  https://policyreview.info/articles/analysis/civil-legal-personality-artificial-intelligence-future-or-utopia, Accessed on 24 November 2021

​

9. ‘Stephen Hawking warns artificial intelligence could end mankind’, Rory Cellan-Jones – URL:  https://www.bbc.com/news/technology-30290540, Accessed on 24 November 2021

​

10. ‘On the road to automated mobility: An EU strategy for mobility of the future’, Communication from the European Commission – URL:  https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52018DC0283, Accessed on 24 November 2021

​

11. ‘The legal framework for autonomous vehicles in the European Union’, Sahin Ardiyok and Armanç Canbeyli – URL: http://www.businessgoing.digital/the-legal-framework-for-autonomous-vehicles-in-the-european-union/, Accessed on 24 November 2021

​

12. ‘Automated vehicles in the EU’, European Parliamentary Research Service – URL: https://www.europarl.europa.eu/RegData/etudes/BRIE/2016/573902/EPRS_BRI(2016)573902_EN.pdf, Accessed on 24 November 2021

​

13. ‘What Are The Different Levels Of Vehicle Automation?’ – URL: https://premioinc.com/blogs/blog/what-are-the-different-levels-of-vehicle-automation, Accessed on 24 November 2021

​

14. ‘Project of the Update of Concept of the Civil Code of Ukraine’ – URL:  https://drive.google.com/file/d/1ExwdnngsmvpAZJtWi836Rr6-x1quaZJQ/view, Accessed on 24 November 2021

​

15. ‘International Human Rights defender on Artificial Intelligence’, Polina Prianykova – URL:  https://prianykovabusiness.wixsite.com/prianykova/post/international-human-rights-defender-on-artificial-intelligence, Accessed on 24 November 2021

​

Officially Published in November 30 - December 03, 2021, Melbourne, Australia

(Table of Contents, № 39)

https://isg-konf.com/wp-content/uploads/2021/11/MODERN-ASPECTS-OF-SCIENCE-AND-PRACTICE.pdf

​

© Copyright by Polina Prianykova_all rights reserved