Artificial Intelligence in Law Enforcement Practice: Risks and Legal Boundaries
Author: Semen Pidluzhniak, Attorney at Grain Law Firm
In modern society, the role of law enforcement lies in the daily work of officers who must direct all their efforts toward creating a safe and fair environment. In such a society, the law becomes a shield for every citizen and shapes a space of security, trust, and legality, where legal decision-making is impossible without achieving a balance between the interests of the state, society, and the individual, including an individual approach to each specific situation while using all available modern methods and technologies.
Since the times of the early civilizations of Babylon, Ancient Egypt, Greece, and Ancient Rome, communities have developed primary mechanisms for protection against external threats and for regulating conflicts within society. Law at that time emerged as a secular phenomenon aimed at resolving disputes, while legal practitioners were creative individuals who assisted in drafting agreements and claims, provided advice in cases of legal uncertainty, and guided parties on how to conduct litigation, relying on their sense of justice.
Today, law enforcement activity constitutes the systematic and continuous activity of state authorities aimed at protecting human rights and freedoms, ensuring public safety, combating crime, and maintaining public order on the principles of legality, professionalism, and respect for human dignity.
Speaking about the phenomenon of integrating artificial intelligence (hereinafter — AI) into law enforcement activities, and considering the modern challenges faced in legal practice, it is important not merely to ask whether AI is useful, but rather to deeply reflect on the key issue: how can we ensure that in every individual case AI does not replace human wisdom, professional experience, independence, and an individualised approach?
Indeed, artificial intelligence is gradually becoming an integral component of the efficiency of law enforcement activities in general. Its use in analytical, investigative, operational-search, and judicial activities significantly improves forecasting accuracy, accelerates the processing of large volumes of data, and automates routine processes. For Ukraine, which is effectively in its fourth year of full-scale war, AI technologies may play a strategic role in combating new forms of cybercrime, identifying hybrid threats such as cyberattacks, manipulation, economic pressure, sabotage, the use of criminal networks, and monitoring critical infrastructure.
Today, there are numerous AI tools capable of facilitating the daily work of legal professionals: document analysis, drafting contracts and claims, client consultations, legal monitoring, and identifying patterns in judicial practice. The use of artificial intelligence in law enforcement practice is no longer surprising; on the contrary, it has become a demand of the times. Educational platforms specialising in legal studies and professional organisations increasingly offer lawyers various training programmes on interacting with AI language models to “simplify” their work. However, in pursuing such “simplification”, legal professionals often become overly creative and increasingly forget the famous phrase of the Roman jurist Ulpian: “law is the art of goodness and justice”, as well as the fact that the primary purpose of law is to create social norms consistent with morality and integrity.
Focusing on the Use of Artificial Intelligence in Justice
Law enforcement authorities are increasingly submitting AI-generated responses as evidence in court or filing motions seeking clarification of court decisions by contrasting judicial reasoning with explanations provided by AI assistants. Process documents also frequently contain definitions of specific concepts or categories generated by artificial intelligence.
Such cases are increasingly reported worldwide, making the issue highly relevant for Ukraine as well. The implementation of technologies in law enforcement, judicial proceedings, and pre-trial investigation is presented as an effective means of modernisation.
Among legal professionals, the following question is being raised: if artificial intelligence already organises routine processes, analyses cases, and formulates decisions, where exactly is the boundary between AI algorithms and human judgment? Whereas the judicial system previously depended entirely on human reasoning, it is now increasingly functioning under the influence of algorithms.
Let us analyse one of the key international standards of justice — Article 6 of the European Convention on Human Rights (ECHR), which guarantees the right to a fair hearing by an independent and impartial tribunal. Although neither this article nor the related commentary contains a direct prohibition on the use of artificial intelligence in justice, the traditional understanding of the administration of justice has always presupposed a professional judge. Such a person delivers judgments based on professional legal education, many years of experience, and individual reasoning, rather than solely on legal information from open sources accessible to AI systems.
Given the current stage of development and capabilities of AI, it may substantially assist judges in handling the heavy caseloads faced by courts worldwide. Increasingly, cases of official use of AI technologies by judges are being recorded in several European countries.
There are also public examples in Ukraine. For instance, during qualification procedures conducted by the High Qualification Commission of Judges of Ukraine (HQCJ) or during the selection of judges for local courts, some candidates openly state that they use artificial intelligence in their practice to draft judicial decisions.
This is further confirmed by the recent Congress of Judges of Ukraine, which approved a new version of the Code of Judicial Ethics. Since September 2024, Article 16 expressly regulates the use of artificial intelligence by judges. Under these amendments, judges may use AI technologies provided that such use does not undermine their independence or impartiality, does not affect the assessment of evidence or the decision-making process, and does not violate legislative requirements.
Indeed, technological progress has simplified the work of legal professionals and appears highly promising. Intelligent systems will continue to improve, learning from our mistakes and adapting accordingly. Artificial intelligence systems may contribute to building a more just and efficient legal system in which decisions are based on objective data and balanced analytical conclusions.
However, the use of artificial intelligence should be permitted exclusively as an auxiliary instrument that simplifies and performs part of the tasks. The absence of unified legal standards governing its use across different jurisdictions creates risks of ethical and regulatory fragmentation, making it necessary to clearly define the limits of its application.
Undoubtedly, the potential of artificial intelligence in law enforcement practice lies primarily in its ability to process vast amounts of data and identify hidden patterns, including those related to criminal activity.
At the same time, there is a danger of excessive automation, where technological solutions begin to displace human judgment, which is unacceptable in a field where moral, ethical, and legal criteria play a decisive role.
Moreover, the number of cases where lawyers are tempted to delegate responses to legal questions to artificial intelligence and excessively rely on AI-generated content is rapidly increasing, potentially undermining the human element of the legal profession.
As a result, there is a risk that legal decisions will be based not on objective legal interpretation and independent legal assessment, but on algorithmic inertia. This creates a dangerous precedent in which technology begins dictating conclusions on legal issues through arbitrary standardisation. It must be remembered that every situation requiring legal analysis should be considered impartially and with regard to its unique circumstances.
Based on my own experience using AI, I can confidently state that chatbots may make mistakes such as incorrect use of sources or mixing information from different responses. In some cases, chatbots combine fabricated details with real personal data, making the information appear credible. Old news may be presented as current, or recent information may be mixed with outdated data. All of this demonstrates how dangerous so-called AI “hallucinations” can be, especially when human rights are involved. Therefore, I always advise verifying references to laws, court decisions, and other primary sources, because responsibility for the accuracy and correctness of any document or advice lies with its author, not with the AI system.
In conclusion, legal professionals should abandon the illusion that artificial intelligence can perform tasks that humans themselves are incapable of performing. Otherwise, a substantial amount of time will ultimately be spent verifying information generated by AI when drafting decisions, requests, or other legal documents.
It is important to shift the focus from the abstract concept of “bias” to the primary source of this problem — “data”. The issue is not that the AI model itself is biased, but rather that it is trained on information that already contains established social narratives.
Considering this risk, one may argue that the use of AI in this context is more compatible with the Anglo-Saxon legal system, which is based on judicial precedent. In such a system, legal principles formulated by judges in previous decisions become binding in resolving similar cases in the future.
However, at the same time, a key feature of the Anglo-Saxon legal system is the leading role of courts in law-making, encompassing both equity and statutory law, which is by no means inherent to AI.
Analysing the importance of ensuring transparency and fairness in the activities of law enforcement and judicial authorities, particularly regarding an individualised approach, it may reasonably be argued that AI technology is incapable of self-analysis or critical reflection. It merely performs tasks based on the data on which it has been trained. Unlike purely logical application of legal norms, law enforcement activity requires profound moral, social, and emotional evaluation of situations, which often becomes decisive for the adoption of a particular decision. Unless mechanisms are implemented to identify and correct biases at the data level, there is a serious risk of automating and legitimising uniformity in resolving similar yet fundamentally different legal issues.
Artificial intelligence is a powerful tool for modernising law enforcement and judicial systems, capable of significantly improving analytical and preventive work. However, its implementation requires caution, regulatory oversight, and public dialogue. Technology must not replace human judgment or undermine the fundamental principles of the rule of law. Properly regulated use of AI may become a guarantee of effective, fair, and transparent law enforcement in which advanced technologies serve people rather than the other way around.
Do We Have Appropriate Legal Regulation of Artificial Intelligence in Ukraine?
No. The rapid implementation of digitalisation processes in the work of law enforcement authorities — including the police, prosecution service, courts, border service, and anti-corruption bodies — is accompanied by numerous challenges, including weak control over algorithms, insufficient protection of data exchange channels, and potential manipulation by third parties. All of this creates conditions for abuse and human rights violations caused by the absence of adequate legal regulation.
The issue of developing a regulatory framework for artificial intelligence technologies is currently being addressed both at the pan-European level and within individual member states. The rapid development of AI technologies has triggered lively discussions within the legal community, making the issue highly relevant. Considerable attention is devoted to this matter in countries such as the United Kingdom, Canada, the United Arab Emirates, and the United States.
Which Path Should Ukraine Follow?
First and foremost, priority should be given to establishing ethical standards for the implementation of artificial intelligence. It is crucial to study European and international experience and, based on best practices, develop legislative regulation of AI-related issues in Ukraine.
Cooperation among all stakeholders — including researchers, developers, leaders of the legal sector, civil society organisations, working groups, and initiatives at all levels — will contribute to harmonising legal policy in this area, preparing proper legislative support, and protecting human rights in the use of artificial intelligence.
And action is required now. Even if Ukraine develops and implements separate national legislation at a certain stage, it will ultimately have to introduce European standards as part of its European integration course, which is enshrined in the Constitution of Ukraine. European experience in overcoming technical difficulties and implementing legal and regulatory standards in this field should serve as a benchmark for Ukraine.
Therefore, before Ukrainian legal professionals proceed to the stage of implementing European legislation, three initial stages may be considered.
The first stage is to inform the public about the advantages and risks of using artificial intelligence. Citizens must be provided with guarantees regarding the protection of their rights and confidentiality when such technologies are used. These reforms should take into account AI standards in light of Ukraine’s current circumstances.
The second stage involves reviewing legislation on personal data protection and human rights to ensure its compatibility with the development of AI technologies and introducing appropriate amendments. It is essential to identify the key legislative gaps and establish rules that encourage responsible innovation and ethical use of AI.
The third stage is to develop a system of principles governing the use of artificial intelligence in law enforcement, including the principles of legality, proportionality, transparency, and human oversight. Every instance of AI application must have a clear legal basis established by law, while interference with private life should be limited to the minimum necessary to achieve a legitimate aim. The final decision in criminal proceedings must remain with a person who understands the legal consequences of their actions and bears responsibility for them.
In this context, it would be appropriate to develop a comprehensive state strategy for the legal regulation of artificial intelligence in the field of justice and law enforcement. Such a strategy should be based on international standards reflected in the recommendations of the Council of Europe, the European Parliament, UNESCO, and the United Nations, all of which emphasise the need for ethical, transparent, and accountable use of technology.
Accordingly, it is critically important for Ukrainian legislation to adapt to these principles by establishing a balance between the efficiency of law enforcement activities and the protection of human rights.
Special attention should also be devoted to the issue of legal liability in cases involving AI errors.
Since artificial intelligence is not a legal entity, responsibility for its actions should rest with specific individuals — developers, operators, law enforcement officers, or judges who decided to use the system. However, Ukrainian legislation currently contains no special rules defining how such liability should be allocated, which creates risks for law enforcement practice and legal certainty.
Lawyers and legislators in Ukraine should actively work on this issue, because the rapid development of AI technologies may significantly displace the human factor in the near future.
At present, memoranda signed between Ukraine and the EU merely establish the parties’ intentions to develop long-term cooperation aimed at involving Ukrainian legal professionals in international processes related to the development of ethical principles and regulation of AI, as well as gaining international practical experience. Therefore, training public officials, lawyers, law enforcement officers, and judges in the use of artificial intelligence, including its ethical aspects, should become one of the country’s priorities.
Legal work in this field should be regulated not only by guidelines on responsible AI use, but also by a comprehensive set of legislative acts. It is critically important that law enforcement decision-making ensures proportionality of legal intervention and avoids both excessive and insufficient state interference. Equally important are equality before the law, predictability of legal consequences, integrity, the ability to manage emotions, avoidance of hasty decisions made under severe stress, and careful consideration of consequences for all parties involved. At the same time, it is necessary to take social norms into account, rely on human intuition, and assume responsibility for the final decision.