Balancing Artificial Intelligence (AI) and Law

Artificial intelligence (AI) is accepted as one of the most important disruptive technologies of our century. The first AI technologies emerged in the 1950s, and with them came a history of dynamic definitions: At first, they were defined as technology that may mirror human intelligence, a concept that persisted until very recently (Kağıtçıoğlu, 2021). By 2021, the European Union coined 'AI' as a software system that predicts, suggests, and decides in place of a human subject (Kağıtçıoğlu, 2021). Today, AI is accepted as the technology that focuses on automating specific types of tasks, especially those that require human intelligence (Surden, 2019). Functions that are special for human beings, such as learning something, sensation, deciding something, or solving a problem, can also be performed by these autonomous systems. However, in spite of these specialties of AI algorithms, most of today’s AI systems do not have a comprehensive and flexible understanding of human beings. Their problem-solving skills are not as developed as those of human beings (Surden, 2019). In brief, whatever its definition is, the most important benefit of AI systems is to support the decision-making mechanisms of humans in complicated and difficult situations (Moslein, 2020).

Figure 1: Use Cases of Artificial Intelligence for the Legal Industry (Trusli, 2021).

New AI systems have started to become visible in several different categories of law, such as due diligence, prediction technology, document automation, intellectual property, legal research, e-billing, and legal analytics (Petch, 2021). They are used to make useful decisions about complex cases, and formal rules can be applied by the algorithms of AI systems (Surden, 2019). Indeed, countries like China have started to use AI judges in their courts for a limited number of subjects already (Kağıtçıoğlu, 2021). On the other hand, there are several ethical questions about AI systems because of the lack of regulations and uncertainties about the functions of these autonomous systems (Petch, 2021). In all conditions, it is certain that this new disruptive technology is going to affect the future legal system in matters of fundamental human rights, administrative law, company law, the law of obligations, and criminal law.

The Relationship Between AI and the Fundamental Rights and Values of People

The first point that needs to be solved by AI technologies is the protection of the fundamental rights and values of human beings. The policy-making process and human rights issues are generally regulated by the constitutions of the states. Fundamental rules about both of these aspects are founded in the constitutions of the states. Nevertheless, the influence of the AI system on the application of the main legal documents of states, such as constitutions, increases every day (Fenaroli, 2022). A state that uses AI algorithms has to minimize the harm to society caused by the destruction of these norms by new autonomous technologies (Fenaroli, 2022). They need to direct their policy-making processes by harmonizing the norms of society with these new disruptive technologies. On the other hand, citizens have to ensure that their fundamental rights in a state’s functions with AI systems are safe. Preventive measures against AI-based discrimination should also be taken (Cognialese, 2021). Indeed, algorithmic-based management can lead to new forms of discrimination in the business world and can generate new concerns about privacy (De Stefano, 2019). As a result, keeping the fundamental values of society and the fundamental rights of their citizens while using AI systems is the obligation of the states. Citizens should be protected from unsafe and ineffective AI systems (Fenaroli, 2022).

Figure 2: The Relationship Between AI and Human Rights (Mpinga, 2021).

The term constitutional AI has been created to prevent the risks related to AI systems (Constitutional AI: The Essential Guide | Nightfall AI Security 101, n.d.).

This new legal term can be explained as "an allocation of rule of law, democracy, and human rights by design" (Nemitz, 2023). Its goal is to safeguard ethical norms when making AI-driven decisions that are effective on human lives, to enforce legal compliance with the legal decisions taken by AI systems, and to enhance public trust in the AI-taken decisions (Surden, 2019). To generate a constitutional AI, constitutional principles must be integrated into the AI training datasets, and a rule-based AI system must be developed (Constitutional AI: The Essential Guide | Nightfall AI Security 101, n.d.).

Another discussion about the relationship between AI technologies and law is the possibility of giving the statute of "person" to autonomous systems. It is vital to define the actions of autonomous systems legally. Indeed, if AI systems are accepted as "persons," they would benefit from the fundamental rights of people. In today’s legal system, non-human actors can also gain the status of "person" if it is necessary for the public good (Kağıtçıoğlu, 2021). The definition of personality in most legal systems depends on behaviors more than appearance, so it is possible that automated AI systems can be accepted as persons in the near future (Willick, n.d.). A developed AI system can have both rights and responsibilities that belong to human beings. The European Council declared that an AI system that can interact with the world without external influence or control must have a legal personality (Kağıtçıoğlu, 2021). However, there are obstacles against this recognition, such as the principle of equality for all men or the different characteristics of men and computers. Today’s computers are far from acting like human beings or being as smart as humans (Willick, n.d.). To solve this contradiction, a legal doctrine claims that they have to be accepted as legal persons, and they need to register with the relevant systems of the administration as companies. On the other hand, there are studies in Europe about a third type of personality in addition to the real person and the legal person: electronic persons (Akbulut & Ag, 2023).

Figure 3: Sophia, First Robot That Appears As A Human (Weaver, 2017).

AI in Administrative Law

Administrative law is defined as the set of rules and principles that govern the relationship between public authorities and civil society (Wolswinkel, 2022). In the future, the administrations of the states may go up on AI algorithms, and most of the administrative issues might be handled with autonomous systems (Cognialese, 2021). In fact, today, the automation process helps states and local authorities manage their public programs and resource allocation strategies (Cognialese, 2021). In the near future, AI technologies could be extremely beneficial and increase the efficiency of the administration of public services (Kağıtçıoğlu, 2021). It is important to detect which fields of administrative law are going to be affected by AI technologies: For administrative issues such as policing, taxation, traffic control, border control, etc., autonomous systems can be extremely beneficial in detecting deficiencies and preventing unlawful actions (Wolswinkel, 2022). States can also benefit from AI technologies for public services related to education, environment, health services, transportation, defense, and energy (Kağıtçıoğlu, 2021). In addition, autonomous systems can also be used to control public servants. By autonomously controlling systems, the misconduct of public servants can be prevented, and their work can be kept within the boundaries of the law (Cognialese, 2021). Another possibility that can be seen in the near future is the interaction of citizens with administrative bodies. Today, this interaction is made among humans. However, in the near future, citizens may interact with AI systems that are created by public authorities (Cognialese, 2021).

Figure 4: A Scheme Which Demonstrate AI Effect on State Administration (Noordt & Misuraca, 2020).

In spite of the existence of all these benefits, the use of AI systems in administrative law is not a simple task. First of all, it will bring a new approach to the responsibility of administrative bodies in administrative law (Kağıtçıoğlu, 2021). For administrative bodies that use AI systems, it will be necessary to explain how their algorithm works and to prove that this algorithm works according to the general principles of law (Cognialese, 2021). Secondly, the equality of arms between public authorities and citizens must also be guaranteed. There should be a limit on the allocation of data according to the data protection law. Irrelevant and unnecessary data about citizens should not be allocated by administrative bodies (Wolswinkel, 2022). In addition, the right of citizens to apply administrative jurisdiction against administrative bodies should not be limited by the usage of these new AI technologies (Wolswinkel, 2022). To be sure about the suitability of AI systems to the general principles of administrative law and the protection of citizens’ rights, there will be a requirement to find new public establishments to develop AI systems that can be used for administrative public issues and to control their functioning. Services that are generated by AI systems can also be controlled by administrative courts and their judges (Kağıtçıoğlu, 2021).

AI in Criminal Law

AI has great potential to be a permanent part of the criminal justice ecosystem; it can be responsible for crimes, a tool to commit a crime, or an assistant for criminal justice professionals (Rigano, 2019). AI systems directed by humans cannot be anything else than a tool that is used to commit a crime. The individual who commits a crime by using an AI system will be accepted as the main criminal of this crime (Lima, 2018). But sometimes, AI systems can commit crimes without human interference. Here, the principles of foreseeability and probability will be effective. The human actor behind the relevant AI algorithm must foresee the consequences of the system's actions. Therefore, in these cases, the creator of the AI algorithm must be accepted as responsible for the criminal action as well (Lima, 2018). In the future, there will be new AI algorithms that will be generated by AI systems instead of humans. For the criminal actions of AI-created AI systems, it is vital to decide whether AI can be accepted as a "person" or not.

Figure 5: An Example of Facial Recognition System in China (AlJazeera, 2023).

In addition to the criminal liability of AI systems, autonomous algorithms can be used as a tool of assistance for current criminal investigations. These systems may be used to identify people, detect medical conditions, and undertake painstaking tasks. Their facial recognition technology will be much more developed. These benefits will be extremely helpful for criminal justice and medical examiners in the future (Rigano, 2019). For complex crimes and accidents, AI algorithms can match faces, identify weapons, and identify other objects that are related to criminal activity. While fulfilling these tasks, they can minimize human errors. They could be a "second pair of eyes" to increase the accuracy of the criminal investigation (Rigano, 2019). All of these benefits are going to increase public confidence in the criminal jurisdiction system and simplify and accelerate the legality and validity of the decision as well (Lapshin, Korneev & Kilimbaev, 2020).

AI in Law of Obligations, Contract Law, and Labor Law

The law of obligations is the basis of all private law fields in all legal systems, and this facet is certainly going to be affected by autonomous systems. The basis of a legal system is constituted by contracts among real persons or legal persons, and these contracts are made by human beings according to the rules of the law of obligations (Soyer & Tettenborn, 2023). However, today’s AI technologies bring the opportunity to write and analyze these contracts without humans (Kağıtçıoğlu, 2021). Especially if there are a high number of contracts that need to be prepared or analyzed, AI systems can easily take care of this job and, if necessary, inform relevant positions such as lawyers or clients (Akbulut & Ag, 2023). This means that for both the universal law of obligations and the universal contract law, a new era will start.

Figure 6: How AI Accelerates the Legal Contract Drafting Process (Vanni, 2020).

The major problem with the AI system for the law of obligations is the detection of concepts such as goodwill and reasonableness (Surden, 2019). If an AI system causes a deficiency or a misconduct by his act, it is uncertain whether it will be the responsibility of the AI system's itself or the responsibility of the real person who wrote the relevant algorithm (Kağıtçıoğlu, 2021). Today, it is impossible for the machine to be responsible for deficiencies, or the payment of compensation to satisfy damages. The creator of the respective AI algorithm will be held responsible for any claims (Soyer & Tettenborn, 2023). However, in the future, there will also be new forms of deficiencies caused by autonomous systems and new treatments for these deficiencies. Without human interactions, AI can cause deficiencies, and it is uncertain whether a computer causing damage counts as the personal fault of a real person or a defect of the machine (Soyer & Tettenborn, 2023). This issue will be related to the acceptance of AI algorithms as "persons" (Kağıtçıoğlu, 2021). As AI continues to develop, this problem will keep growing (Soyer & Tettenborn, 2023).

The relationship between labor law and AI systems demonstrates itself in the discussions about how AI systems will affect the jobs of the future. It is as important as the implementation of autonomous systems in the law of obligations. There is a general belief about the probability that a particular job in our world is going to be automated by AI technologies, which will cause a wave of unemployment around the world. There will be an impact of new technologies in the 21st century on the workforce. As a result, several monotonous and hazardous jobs are going to disappear (De Stefano, 2019). It will be the main challenge of labor law around the world to determine how the rights of laborers will be protected during this automation process and how the introduction of these technologies into business life will be governed (De Stefano, 2019).

Figure 7: Jobs That Will Automated For People, Until 2030 (Economist, 2018).

AI in Commercial Law and Corporate Law

Commercial law and corporate law (company law) are fields of law about trade, finance, commerce, and competition. Companies have their own personalities, and based on the actions of companies, the board is accepted as the responsible organ (Petch, 2021). Today’s general principles of company law are determined by human executives; the decision-making process can be extremely complicated for these people (Moslein, 2020). Especially for decisions that have to be taken by using a huge amount of data, robo-executives can take more beneficial and certain decisions for companies (Moslein, 2020). So, in the following years, robo-executives and robo-board members may be effective in company administrations (Akbulut and Ag, 2023). For that reason, it is undeniable that the commercial sector and its legal aspects are going to be affected a lot by AI systems. There will be new challenges for corporate law and commercial law systems around the world (Moslein, 2020).

Another challenge for company law may be accepted as the delegation issue in a company. The members of the boards of a company can easily delegate their duties to the managers, and under this condition, they need to control and instruct these delegated managers (Moslein, 2020). Both of these necessities can easily be given to the AI systems. On the other hand, an AI system can be delegated by a member of the board as well. The suitability of this condition for the general principles of corporate law is uncertain (Moslein, 2020).

Figure 8: First Robot CEO in the World, MIKA (Dixit, 2019).

Conclusion

One of the most important barristers in the United Kingdom, Robin Allen QC, claimed that AI is going to change the world for the better, but if used in the wrong way, it would represent an exceptionally dangerous tool for humanity (Petch, 2021). We can accept this statement as a good summary of the future of autonomous technologies. By using AI systems, law systems and legal rules might be more understandable, manageable, useful, accessible, and predictable (Surden, 2019). However, if social norms are harmed by this technology or if the statute of what constitutes a "person" continues to be uncertain, it will be a threat to the future of humanity. Addressing the current gaps in legal scholarship and codes may aid the fruitful integration of artificial intelligence in its various roles.

Bibliographical References

Akbulut, Gizem, & Ag, Ahmet. (2023). Innovative Applications of Artificial Intelligence in Commercial Law. GSI Articletter, 28, 40-59.

Coglianese, C. (2021). Administrative Law in The Automated State. Daedalus, 150(3), 104-120.

Constitutional AI: The Essential Guide | Nightfall AI Security 101. (n.d.). https://www.nightfall.ai/ai-security-101/constitutional-ai

De Stefano, Valerio. (2019). Automation, Artificial Intelligence, and Labor Protection. Comparative Labor Law & Policy Journal, 41(1), 3-14.

Fenaroli, F. (2022, December 21). Rights and obligations of Artificial Intelligence: where does the law stand? Neosperience. https://www.neosperience.com/blog/rights-and-obligations-of-artificial-intelligence-where-does-the-law-stand/

Kagitcioglu, Mutlu. (2021). Artificial Intelligence and Administrative Law (An Assessment From Today to the Future). Hacettepe Law Faculty Magazine, 11(1), 118-168.

Lapshin, V. F., Korneev, S. A., & Kilimbaev, R. V. (2020, December). The Use Of Artificial Intelligence In Criminal Law and Criminal Procedure Systems. In IOP Conference Series: Materials Science and Engineering (Vol. 1001, No. 1, p. 012144). IOP Publishing.

Lima, Dafni. (2018). Could Ai Agents Be Held Criminally Liable: Artificial Intelligence and The Challenges For Criminal Law. South Carolina Law Review, 69(3), 677-696.

Nemitz, P. (2019). Constitutional Democracy and Technology In The Age Of Artificial Intelligence. The Royal Society Publishing.

Moslein, F. (2021). Robots In The Boardroom: Artificial Intelligence And Corporate Law. Istanbul University Faculty of Law Magazine, 79(2), 699-728. https://doi.org/10.26650/mecmua.2021.79.2.0004

Petch, J. (2021, September 26). Artificial Intelligence in Commercial Law: Legal Evolution or Hindrance? Oscola. https://www.oscola.org/post/artificial-intelligence-in-commercial-law-legal-evolution-or-hindrance-jenny-petch 

Rigano, C. (2019). Using Artificial Intelligence to Address Criminal Justice Needs. NIJ Journal, (280).

Soyer, B., & Tettenborn, A. (2023). Artificial Intelligence and Civil Liability—Do We Need A New Regime? International Journal of Law and Information Technology, Oxford Press, 30, 385–397. https://doi.org/10.1093/ijlit/eaad001 

Surden, H. (2018). Artificial Intelligence and Law: An Overview. Georgia State University Law Review, 35(4).

Willick, M. S. (1985, August). Constitutional Law and Artificial Intelligence: The Potential Legal Recognition of Computers as" Persons". In IJCAI (pp. 1271-1273).

Wolswinkel, J. (2021). Artificial Intelligence and Administrative Law, Comparative Study On Administrative Law And The Use Of Artificial Intelligence And Other Algorithmic Systems In Administrative Decision-Making In The Member States Of The Council Of Europe. Council of Europe.

Visual Sources