Artificial Intelligence as a New Source of Victimization: Criminal Law and Civil Law Protection of Victims

о

Modern technological developments, espe‑ cially in the field of artificial intelligence, have led to profound changes in the way society, the economy and the legal system function. Artificial intelligence is now being applied in numerous areas, including healthcare, the fi‑ nancial sector, the judiciary, security systems, autonomous vehicles and digital platforms. Although these technologies contribute to in‑ creasing efficiency, automating processes and improving the quality of life, their application simultaneously creates new risks to the rights, security and legal position of individuals. Of particular importance is the fact that artificial intelligence systems can autonomously make decisions that produce legally relevant con‑ sequences, including the occurrence of dam‑ age and the infringement of legally protected goods.

In this context, artificial intelligence ap‑ pears as a new source of victimization, because its application can lead to various forms of damage, including property damage, infringe‑ ment of personal rights, discrimination, viola‑ tion of privacy, but also endangering the life and physical integrity of individuals. Victim‑ ization caused by artificial intelligence can occur as a result of technical errors, system design flaws, inadequate oversight, and in‑ tentional misuse of these technologies by humans. A particular problem is the fact that the autonomy and complexity of artificial in‑ telligence systems make it difficult to establish a causal link between an action and the result‑ ing damage, as well as to identify the responsi‑ ble entity, which can lead to legal uncertainty and make it difficult to protect victims.

Traditional legal systems are based on the assumption that the bearers of legal respon‑ sibility are natural and legal persons who act with a certain degree of awareness and will. However, autonomous artificial intelligence systems can operate without direct human control, which leads to the emergence of the so‑called criminal and civil liability gap, i. e. situations in which damage occurs but it is difficult to determine the responsible party. This raises fundamental questions about the application of existing legal institutions, in‑ cluding criminal liability, civil liability, pro‑ ducer liability, liability for dangerous goods and activities, and liability of legal persons.

Criminal law plays a key role in protect‑ ing fundamental social values and individ‑ uals from the most serious forms of threat, including criminal offenses committed using artificial intelligence. In such cases, criminal liability cannot be attributed to the artificial intelligence system itself, but is transferred to the persons who participated in its deve‑ lopment, implementation or use, if they ac‑ ted with intent or negligence. On the other hand, civil law provides protection for victims through the institute of compensation for damage, enabling injured parties to obtain fair compensation for material and non‑material damage suffered. Civil liability is of particular importance because it enables protection of victims even in situations where there are no elements of criminal liability, but there is an obligation to compensate for damage.

Based on the above, the subject of this pa‑ per is the analysis of artificial intelligence as a new source of victimization, with a special focus on the criminal and civil protection of victims. The aim of the paper is to examine whether existing legal institutions provide adequate protection for victims of damage caused by artificial intelligence, as well as to identify key challenges and directions for fur‑ ther development of the legal framework in this area. Special attention will be paid to the analysis of the bases of legal liability, identifi‑ cation of responsible entities and mechanisms of legal protection, with the aim of determin‑ ing the efficiency of the existing system and the need for its improvement in conditions of accelerated technological development.

Artificial Intelligence as a New Source of Victimization

The development and application of artifi‑ cial intelligence have led to the emergence of new forms of victimization that differ from traditional sources of harm both in the way they arise and in the complexity of determin‑ ing responsibility. Unlike classical technical means, artificial intelligence systems have the ability to make autonomous decisions, process large amounts of data and adapt their own behavior based on previous experience, which increases the risk of harmful consequences for individuals. It is precisely this autonomy that is the key characteristic that makes artificial intelligence a specific source of victimization, because it allows for decisions that can have legally relevant consequences without direct human intervention [1, p. 167–169].

Artificial intelligence can cause various forms of harm, including physical, proper‑ ty, psychological and reputational damage. A special problem is the fact that artificial intelligence systems can make decisions that directly affect the rights and legal status of individuals, including decisions in the areas of employment, financial services, health care and security. In such cases, damage may oc‑ cur as a result of a technical error, inadequate system design or unpredictable algorithmic behavior, which creates a need for legal pro‑ tection of victims [2, p. 2–4].

An additional problem is the lack of trans‑ parency of algorithmic decision‑making, often referred to as the “black box” phenomenon, which makes it difficult to establish a causal link between the operation of an artificial intelligence system and the damage caused, which further complicates the possibility of proving liability and achieving legal protec‑ tion for victims [1, p. 170–172; 3, p. 433–435]. Artificial intelligence can also be a means of committing criminal acts, which directly in‑ creases the scope of victimization. Criminals are increasingly using artificial intelligence to commit various forms of cybercrime, including fraud, data manipulation and identity theft, with these technologies enabling the commis‑ sion of criminal acts on a larger scale and with a higher degree of efficiency than before. The use of artificial intelligence allows criminals to automate attacks and simultaneously victim‑ ize a large number of people, which increases the social danger of these forms of crime [4, p. 77113–77116].

In addition to property and digital dam‑ age, AI can also cause physical harm, espe‑ cially in the case of autonomous systems such as autonomous vehicles, medical robots or industrial systems. In such situations, the autonomous system can cause bodily harm or death, making AI the direct cause of victim‑ ization. Such cases raise the issue of attribu‑ tion of liability, since AI does not have legal personality or the ability to navigate in a legal sense, which means that liability must be at‑ tributed to the people or legal entities behind its development or application [5, p. 6–8].

A special form of victimization is the violation of personal rights caused by AI, including violations of the right to privacy, discrimination and damage to reputation. Automated decision‑making systems can lead to discriminatory decisions or unjusti‑ fied restrictions on the rights of individuals, which is a basis for civil law protection. In such cases, victims have the right to claim compensation for damages and protection of their personal rights, which confirms the im‑ portance of civil law protection mechanisms [6, p. 45–49].

Artificial intelligence can cause victim‑ ization even in situations where there is no intention to cause damage, but damage occurs as a result of a technical defect or unpredict‑ able behavior of the system. Such cases are particularly significant because they show that victimization can occur even without unlawful intent, which further complicates the appli‑ cation of traditional legal institutions based on fault. Therefore, modern legal theory in‑ creasingly emphasizes the need to apply strict liability as a mechanism for protecting victims of damage caused by artificial intelligence [7, p. 13–15].

The specificity of victimization caused by artificial intelligence is also reflected in the complexity of the relationship between differ‑ ent entities participating in its development and application. The process of developing and applying artificial intelligence involves the developer, manufacturer, operator and user of the system, which makes it difficult to determine responsibility and can lead to the so‑called “multiple participants” problem, i. e. a situation in which it is difficult to identify the responsible person. This complexity fur‑ ther increases the need to develop an adequate legal framework that will ensure effective pro‑ tection of victims [8, p. 1073–1077].

Based on the above, it can be concluded that artificial intelligence represents a qual‑ itatively new source of victimization, which differs from traditional sources of harm in its autonomy, complexity and decision‑making ability. These characteristics create new legal challenges and require the adaptation of ex‑ isting legal institutions in order to ensure ad‑ equate protection of victims. For this reason, the analysis of criminal and civil mechanisms for the protection of victims of artificial intel‑ ligence represents one of the key issues of the modern legal system.

Criminal Protection of Victims of ArtificialIntelligence

Criminal law plays a key role in protecting fundamental social values and individuals from the most serious forms of violations of legally protected goods, including damage caused by the use of artificial intelligence. With the development of autonomous sys‑ tems, new forms of criminal behavior appear in which artificial intelligence can be a means of committing a crime or a direct cause of harmful consequences. These circumstances require the application of existing criminal law institutes, but also their adaptation to new technological conditions in order to ensure adequate protection of victims [5, p. 4–6].

One of the basic issues of criminal law protection of victims of artificial intelligence relates to the possibility of attributing crimi‑ nal liability. The traditional concept of crim‑ inal liability is based on the existence of an unlawful act (actus reus) and guilt (mens rea), i. e. the consciousness and will of the perpe‑ trator. However, artificial intelligence does not possess consciousness, intention or the ability to understand the consequences of its actions, which is why it cannot be the bearer of crim‑ inal liability in the legal sense [9, p. 37]. This characteristic means that criminal liability cannot be attributed to the artificial intelli‑ gence system itself, but rather the liability of natural or legal persons who participated in its development, implementation or use must be established.

In this sense, criminal liability can be at‑ tributed to developers, manufacturers, opera‑ tors or users of artificial intelligence systems, if their actions contributed to the occurrence of damage. Developer liability may exist in the case of inadequate system design, testing failures or conscious ignoring of security risks, while operator liability may exist in the case of inadequate supervision or improper use of the system [10, p. 5–8]. Also, legal persons may bear criminal liability if the damage occurred as a result of organizational failures or lack of adequate security measures [8, pp. 1076– 1079].

Of particular importance is the insti‑ tute of the so‑called indirect commission of a criminal offense, according to which artifi‑ cial intelligence can be viewed as a means of committing a criminal offense, while liability lies with the person who uses or controls such a system. In such cases, artificial intelligence is an instrument used to commit an illegal act, while criminal liability is attributed to the per‑ son who enabled or predicted such behavior of the system [5, p. 9–11]. This model enables the application of existing criminal law norms and ensures the protection of victims in situations where artificial intelligence is used as a means of committing a criminal offense.

Criminal liability can also exist in cases of negligence, i. e. when the responsible per‑ son has not taken the necessary measures to prevent the occurrence of damage. If the developer, manufacturer or operator has not ensured an adequate level of system security, although he could and should have foreseen the possibility of damage, there may be crim‑ inal liability for the consequences caused by the operation of the artificial intelligence sys‑ tem [9, p. 138–142]. This form of liability is of particular importance for the protection of victims, because it allows the determination of criminal liability even in cases where there is no intent, but there is an omission in the action.

One of the greatest challenges in the criminal law protection of victims is the so‑ called criminal liability gap, i. e. a situation in which artificial intelligence causes damage, but no natural or legal person meets all the elements of criminal liability. Such situations can arise due to the autonomy of the system, the complexity of algorithms and the inability to predict all possible consequences of their work [11, p. 443–447]. This problem can lead to a situation in which the victim is left with‑ out adequate criminal law protection, which indicates the need for further development of the legal framework in this area.

Artificial intelligence can also be used to commit various forms of cybercrime, includ‑ ing fraud, identity theft, data manipulation and other forms of criminal behavior. The use of artificial intelligence enables the commis‑ sion of criminal offenses on a larger scale, with a higher degree of efficiency and a lower risk of detection of the perpetrator, thereby in‑ creasing the number of victims and the social danger of these offenses [4, p. 77114–77118]. In such cases, criminal law protection of vic‑ tims is achieved through the identification and punishment of responsible persons, as well as through the application of appropriate crim‑ inal sanctions.

Criminal law also has a preventive func‑ tion, as the application of criminal sanctions contributes to the prevention of future harm‑ ful actions and the protection of potential victims. Establishing criminal liability of de‑ velopers, manufacturers and users of artificial intelligence systems contributes to increasing accountability and encourages the implemen‑ tation of higher safety standards, thereby re‑ ducing the risk of harm [10, p. 8–10]. In this way, criminal law not only ensures the pro‑ tection of existing victims, but also contrib‑ utes to the prevention of future victimization. Based on the above, it can be concluded that criminal law is an important mechanism for protecting victims of harm caused by artificial intelligence, although there are certain chal‑ lenges in its implementation. The inability to attribute criminal liability to the artificial intelligence system itself, as well as the com‑ plexity of determining the responsibilities of various participants in its development and implementation, require the adaptation of existing legal institutions. Despite these challenges, criminal law remains a key instru‑ ment for protecting victims and ensuring legal certainty in the conditions of the rapid devel‑ opment of artificial intelligence.

Civil Law Protection of Victims of ArtificialIntelligence

Civil law liability is the basic mechanism for protecting victims of damage caused by arti‑ ficial intelligence, as it enables the exercise of the right to compensation for damage and the establishment of a fair balance between the injured party and the responsible party. Unlike criminal law, whose primary function is to punish the perpetrator, civil law has a com‑ pensatory and corrective function, aimed at protecting the interests of the victim and eliminating the consequences of the damage. Civil law liability implies the obligation of the responsible entity to compensate for the dam‑ age that occurred as a result of its actions or the actions of the system under its control, thereby ensuring the legal protection of the injured party [1, p. 166–168].

One of the main problems of civil law protection of victims of artificial intelligence relates to the fact that autonomous systems can make decisions without direct human in‑ tervention, which makes it difficult to deter‑ mine liability. Despite this, modern legal the‑ ory assumes that liability must be attributed to entities that participate in the development, implementation or use of artificial intelligence systems, including developers, manufacturers, operators and legal entities. These entities are liable for damage caused by the operation of the system, because they have control over its functioning and the ability to prevent harmful consequences [2, p. 2–5].

Civil liability can be based on various grounds, including subjective liability based on fault and objective liability based on risk. Subjective liability exists when the damage occurred as a result of the omission, negli‑ gence or inadequate conduct of the respon‑ sible person, while objective liability exists regardless of fault, when the damage oc‑ curs as a result of the action of a dangerous thing or activity. Given the complexity and potential danger of artificial intelligence systems, modern legal theory increasingly emphasizes the importance of objective lia‑ bility as an effective mechanism for protect‑ ing victims, as it allows for compensation for damage even when it is not possible to prove the guilt of the responsible person [3, p. 434–438].

Of particular importance is liability for things and machines, which can also be ap‑ plied to artificial intelligence systems. In such cases, the person who has control over the sys‑ tem, i. e. its owner or operator, is liable, as he has an obligation to ensure its safe use and prevent damage. This principle enables the protection of victims even in situations where damage occurs as a result of the autonomous functioning of the system, because liability arises from the very fact of control over the system [7, p. 14–16].

Civil liability is of particular importance in cases where artificial intelligence causes a vio‑ lation of personal rights, including the right to privacy, the right to equality and the right to protection of the person. Automated decision‑ making systems can lead to discrimination, damage to reputation or unlawful processing of personal data, which creates the basis for civil protection of victims. In such cases, the injured party has the right to claim compen‑ sation for material and non‑material damage, as well as the cessation of further violations of rights [6, p. 46–50].

One of the most important forms of civ‑ il protection of victims is the liability of the manufacturer for a defective product. If an artificial intelligence system contains a tech‑ nical defect or is not adequately designed, the manufacturer may be liable for damage caused by its use. This liability is of particular importance because it encourages manufac‑ turers to apply high standards of safety and quality, thereby reducing the risk of damage and contributing to the protection of victims [3, p. 437–440].

Civil protection of victims can also be achieved through the liability of developers and operators of artificial intelligence systems. The developer may be liable if the system is not adequately designed or tested, while the oper‑ ator may be liable if he or she has not provided adequate supervision over the system. This liability is based on the fact that these persons have the ability to control the system and the obligation to prevent damage [12, p. 61–66].

The role of insurance as a mechanism for protecting victims is also of particular impor‑ tance. The introduction of mandatory insur‑ ance for artificial intelligence systems allows for the provision of compensation to victims even in situations where the responsible per‑ son is unable to compensate for the damage. Insurance contributes to more efficient pro‑ tection of victims and reduces the risk of the injured party being left without compensation [2, p. 6–7].

Civil liability also has a preventive func‑ tion, as it encourages developers, manu‑ facturers and users of artificial intelligence systems to implement appropriate security measures to prevent damage. The existence of the possibility of civil liability contributes to increasing the responsibility of participants in the development and application of artifi‑ cial intelligence, thereby reducing the risk of victimization [6, p. 51–53].

The analysis conducted indicates that civil law is a key mechanism for protecting victims of damage caused by artificial intelligence. Through the institutes of compensation for damage, strict liability, producer liability and liability for dangerous goods, civil law enables effective protection of victims and ensures fair compensation for the damage caused. Al‑ though the autonomy and complexity of arti‑ ficial intelligence systems create certain chal‑ lenges in the application of these institutes, civil liability remains the main instrument for protecting victims and ensuring legal certainty in a modern digital society.

Challenges and Directions for improving Legal Protection of Victims of Artificial Intelligence

Despite the existence of criminal and civil protection mechanisms, the application of these institutes in the context of artificial intelligence faces numerous challenges. One of the main problems is the autonomy of the artificial intelligence system, which makes it difficult to establish a causal link between the system’s actions and the resulting damage, as well as to identify the responsible entity. Given that artificial intelligence systems can make decisions without direct human control, it is often difficult to determine whether the developer, manufacturer, operator or user of the system is responsible, which can lead to legal uncertainty and difficult protection of victims [1, p. 170–172; 8, p. 1073–1077].

A special challenge is the so‑called liabil‑ ity gap, i. e. a situation in which an artificial intelligence system causes damage, but it is not possible to establish the criminal liability of a specific person. This problem arises from the fact that artificial intelligence does not have legal personality or the ability to reason and make decisions in a legal sense, which is why liability must be attributed to the people or legal entities behind its development and application [11, p. 443–447; 9, p. 134–137]. In such situations, the criminal protection of victims may be limited, which further empha‑ sizes the importance of civil law protection mechanisms.

An additional problem is the complexity and non‑transparency of artificial intelligence systems, which makes it difficult to prove li‑ ability. Complex algorithms and the lack of transparency in the decision‑making process can make it difficult to prove error or omission, thereby reducing the victim’s ability to obtain legal protection [10, p. 6–9]. These problems point to certain limitations of traditional legal institutions that have been developed in the context of human action and technical sys‑ tems with predictable behavior. Therefore, modern technological development imposes the need to review the existing legal concepts of liability and adapt them to the conditions of the digital society.

In this sense, modern legal theory in‑ creasingly emphasizes the need to develop a special regulatory framework for artificial intelligence, which would clearly define the liability of all participants in its development and application. Of particular importance is the introduction of strict liability and man‑ datory insurance, as well as a more precise definition of the obligations of developers, manufacturers and operators of artificial intelligence systems [2, p. 6–8; 3, p. 439–442]. Such solutions would contribute to more ef‑ ficient protection of victims and reduction of legal uncertainty, especially in situations where proving guilt is a significant obstacle to obtaining compensation for damages.

Based on the above challenges, it is possi‑ ble to identify several directions for the devel‑ opment of legal regulations in this area. First of all, legislators could consider introducing a special strict liability regime for high‑risk artificial intelligence systems, based on the risk principle, which would facilitate the ob‑ taining of compensation for damages in situ‑ ations where it is difficult to prove the guilt of a specific entity. In addition, the normative framework should more precisely define the liability of all participants in the life cycle of artificial intelligence systems, including de‑ velopers, manufacturers, operators and users of the system. A clear demarcation of their obligations and responsibilities could contrib‑ ute to reducing the so‑called liability gap and provide greater legal certainty.

An additional protection mechanism could be a mandatory insurance model for certain categories of autonomous and high‑risk arti‑ ficial intelligence systems. The introduction of such a model would enable more efficient and faster compensation for damages to vic‑ tims, even in situations where the responsible party is unable to compensate for the damage caused. At the same time, mandatory insur‑ ance could contribute to increasing safety standards in the development and application of these technologies.

Finally, civil law mechanisms are of par‑ ticular importance in protecting victims of artificial intelligence, as they enable the real‑ ization of compensation for damages even in situations where criminal liability cannot be established. Civil law, as a corrective and pre‑ ventive mechanism, contributes to the protec‑ tion of individual rights and the provision of fair compensation for the damage caused, thus representing a key instrument of legal protec‑ tion in a modern digital society [6, p. 51–54].

Conclusion

The development and widespread application of artificial intelligence have led to the emer‑ gence of new forms of victimization that pose a significant challenge to modern legal sys‑ tems. The specificity of these forms of victim‑ ization is reflected in the autonomy of artifi‑ cial intelligence systems, their ability to make decisions without direct human control, as well as the complexity of establishing a causal link and legal liability for the damage caused. Artificial intelligence can cause various forms of damage, including property, physical, psy‑ chological and reputational damage, which can put individuals in the position of victims, which requires adequate and effective legal protection.

The analysis shows that criminal law plays an important role in protecting victims of damage caused by artificial intelligence, as it allows for the determination of the liability of natural and legal persons behind the de‑ velopment, implementation or use of these systems. However, given that artificial intel‑ ligence does not have legal personality or the ability to reason, criminal liability cannot be attributed to the system itself, but rather the liability of the developer, manufacturer, opera‑ tor or user must be determined, if their actions contributed to the damage. Although this ap‑ proach provides a certain level of protection, the autonomy and complexity of artificial in‑ telligence systems can in some situations lead to the so‑called liability gap, which indicates certain limitations of criminal law protection.

Civil law is a key mechanism for protecting victims, as it enables the exercise of the right to compensation for damage and the estab‑ lishment of a fair balance between the injured party and the responsible party. Through the institutes of subjective and objective liability, producer liability and liability for dangerous goods and activities, civil law enables effective compensation to victims even in situations where criminal liability cannot be established. Therefore, civil liability has a particularly im‑ portant compensatory and preventive function in the context of damage caused by artificial intelligence.

However, the rapid technological develop‑ ment and the increasing autonomy of artifi‑ cial intelligence systems indicate the need for further development of the legal framework in this area. In this regard, modern legal systems should consider introducing special regulatory mechanisms that would include a clearer defi‑ nition of the responsibilities of all participants in the development and implementation of ar‑ tificial intelligence systems, as well as a wider application of objective liability for high‑risk artificial intelligence systems. In addition, the introduction of mandatory insurance for cer‑ tain categories of autonomous systems could contribute to more efficient and faster com‑ pensation of damages to victims.

Further development of legal regulations in the field of artificial intelligence must be aimed at establishing a balance between en‑ couraging technological innovations and en‑ suring effective legal protection of individuals. Only through clearly defined mechanisms of responsibility, transparency in the operation of artificial intelligence systems and adequate compensation mechanisms is it possible to ensure effective protection of victims and preserve legal certainty in a modern digital society.