Artificial intelligence in Ukrainian traditional categories of criminal law

Размещено на http://www.allbest.ru/

Luhansk State University of Internal Affairs named after E.O. Didorenko

Yaroslav Mudryi National Law University

Mykola Karchevskyi, Doctor of Law, Professor, Vice-rector, Member of Ukrainian National Group of AIDP-IAPL

Oleksandr Radutniy, PhD in Law, Associate Professor (Docent) of the Criminal Law Department, Member of Ukrainian National Group of AIDP-IAPL

At the request of the International Association of Criminal Law (AIDP- IAPL, Association International de Droit Penal - a non-governmental organization on criminal law, Paris, France) within the framework of the XXI International Congress of Criminal Law ”Artificial Intelligence and Criminal Justice”, a subgroup of the Ukrainian national group of AIDP-IAPL in two scientists prepared detailed answers to questions about * definition and legal qualification of Artificial Intelligence system (legal definition of AI system in Ukrainian law, Ukrainian AI-based systems for predictive policing, legal definition of Machine Learning in Ukrainian law, legal personhood or legal capacity to the AI systems), * existing criminal offences and criminalization (illegal act committed by, through or against an AI system; new offences related to designing, programming, developing, producing, functioning or making use of AI systems; crimes of mere conduct, commission and omission offences, consummate offence, crimes with intent, etc.; who can be considered the possible perpetrator and/or victim of the new AI offences e.g. producers / programmers / system engineers / developers / designers etc.; specific mental element of individual criminal liability, possibility for legal persons be held liable for AI crimes committed by any person acting individually or having a leading position within the legal person; criminal responsibility of the perpetrator or of the legal person in order to avoid the risk of over- criminalization if the AI systems are produced, used or put on the market for legal purposes, e.g. for scientific or research reason; whether reports or legal literature suggest the introduction of new criminal offences linked to AI systems; positive obligations for persons and/or legal person designing, developing, producing, testing, selling or distributing AI systems etc.), * applicability of Traditional Criminal Law Categories (possibility for AI system considered as a “computer system ” as defined by Article 1, lett. of Cybercrime Convention and or Article 2, lett. a) of Directive EU/2013/40; specific problems with respect to the principle of legality; admissibility of analogy in order to criminalize illegal acts related to AI systems; joint-perpetrator or participant in the commission of the crime; forms of secondary liability applicable to AI-related crimes; state of mind (e.g. dolus) on the part of the human agent who designed / programmed / developed / produced / circulated / marketed / used the AI system; exact and concrete modus operandi of the AI system in committing the offence etc.), * adaptation of Traditional Criminal Law Categories and academic debate (principle of culpability nullum crimen sine culpa and mens rea; compliance with the principle of culpability when the output causing the harm generated by the intelligent machine is neither wanted nor predictable by the human agent; compliance with the principle of culpability when an AI system is intentionally used by a human agent as a tool but the AI system carried out an offence different from the one wanted by the human agent; criminal participation and attempted crimes; liability of legal persons; necessary adjustments of the legal principles on criminal liability of legal persons when they are involved in AI-related crimes; necessary adjustments of policies and preventive measures within private organizations in order to guarantee a correct and regular use of AI systems etc.), * alternatives to criminalization and non-criminal sources.

Key words: artificial intelligence, criminal liability, criminal law, criminal justice, criminal offense, Association International de Droit Penal, qualification, machine learning, legal personality of artificial intelligence, legal capacity of artificial intelligence, criminalization, excessive criminalization, perpetrator, victim person, responsibility of legal entities, positive obligations, principle of legality, principle of culpability, analogy, state of mind (dolus) of human agent, nullum crimen sine culpa, mens rea.

Карчевський М. В., Радутний О. Е. Штучний інтелект в традиційних категоріях кримінального права України

На прохання Міжнародної асоціації кримінального права (AIDP-IAPL, Association International de Droit Penal - неурядової організації з кримінального права, Париж, Франція) у рамках XXI Міжнародного конгресу кримінального права «Штучний інтелект і кримінальне правосуддя» підгрупа Української національної групи AIDP-IAPL у складі двох науковців підготувала розгорнуті відповіді на питання щодо * визначення та правової кваліфікації штучного інтелекту (юридичне визначення штучного інтелекту в українському законодавстві, огляд національних систем штучного інтелекту для інтелектуальної поліцейської діяльності, правове визначення машинного навчання в українському законодавстві, правосуб'єктність або правоздатність систем штучного інтелекту тощо), * наявних кримінальних правопорушень та їх криміналізації (незаконні дії, вчинені системою штучного інтелекту за допомогою системи штучного інтелекту або проти неї; нові правопорушення, пов'язані з проєктуванням, програмуванням, розробкою, виробництвом, функціонуванням або використанням систем штучного інтелекту; злочини, пов'язані з простою поведінкою; персона виконавця та/або жертви нових правопорушень, зокрема, виробники / програмісти / системні інженери / розробники / дизайнери тощо; специфічний психічний елемент індивідуальної кримінальної відповідальності, можливість для юридичних осіб нести відповідальність за злочини зі штучним інтелектом, вчинені будь-якою особою, яка діє індивідуально або займає керівну посаду в юридичній особі; межі кримінальної відповідальності злочинця або юридичної особи з огляду на уникнення ризику надмірної криміналізації, якщо системи штучного інтелекту виробляються, використовуються або пропонуються на ринок для законних цілей, наприклад, для наукових або дослідницьких цілей; огляд звітів або юридичної літератури щодо пропозицій впровадження нових кримінальних злочинів, пов'язаних із системами штучного інтелекту; позитивні зобов'язання для осіб та/або юридичних осіб, які проєктують, розробляють, виробляють, тестують, продають або розповсюджують системи штучного інтелекту тощо), * можливості застосування категорій традиційного кримінального права (доцільність розуміння системи штучного інтелекту як «комп'ютерної системи» згідно з положеннями статті 1 Конвенції про кіберзлочинність та/або пунктом «a» статті 2 Директиви EU/2013/40; специфічні проблеми дотримання принципу законності; допустимість аналогії для криміналізації протиправних дій, пов'язаних із системами штучного інтелекту; форми вторинної відповідальності до злочинів, пов'язаних із штучним інтелектом; стан розуму (наприклад, dolus) з боку агента-людини, який проєктував / запрограмував / розробив / виготовив / поширив / продав / використав систему штучного інтелекту; точний і конкретний спосіб дії системи штучного інтелекту при вчиненні правопорушення тощо), * адаптації категорій традиційного кримінального права та наукових дебатів (принцип винності nullum crimen sine culpa і mens rea; його дотримання, коли суспільно небезпечний результат заподіяний розумною машиною, не є ані бажаним, ані передбачуваним для людини; відповідність принципу винної відповідальності, коли система штучного інтелекту навмисно використовується агентом-людиною як інструмент, але система штучного інтелекту вчинила злочин, відмінний від того, якого бажає агент-людина; необхідність коригування правових принципів кримінальної відповідальності юридичних осіб, якщо вони причетні до злочинів, пов'язаних із штучним інтелектом; необхідність коригування політики та превентивних заходів у приватних організаціях для гарантування правильного та регулярного використання систем штучного інтелекту тощо), * альтернатив криміналізації та некримінальних джерел відповідальності.

Ключові слова: штучний інтелект, кримінальна відповідальність, кримінальне право, кримінальне правосуддя, кримінальне правопорушення, Association International de Droit Penal, кваліфікація, машинне навчання, правосуб'єктність штучного інтелекту, правоздатність штучного інтелекту, криміналізація, надмірна криміналізація, виконавець, потерпіла особа, відповідальність юридичних осіб, позитивні зобов'язання, принцип законності, принцип винної відповідальності, аналогія, стан розуму (dolus) агента-людини, nullum crimen sine culpa, mens rea.

Introduction

Lately the Ukrainian National Group of the International Association of Penal Law (AIDP-IAPL, Association International de Droit Penal - a criminal law non-governmental organization located in Paris, France) received the task of preparing the report “Traditional Criminal Law Categories and AI: Crisis or Palingenesis?” for the International Colloquium of Section I (Criminal Law - general part) of the XXI^th International Congress of Penal Law “Artificial Intelligence and Criminal Justice”. The research in Section I was conducted by the authors of this article. The task consisted in answering pre-formulated questions. Some of them are offered to your attention in the following form. Due to the constant development of the legislation and the improvement of the scientific views of the authors, the given answers may be clarified.

1. Definition and legal qualification of Artificial Intelligence system

1. Is there a legal definition of AI system in Ukrainian law? On September 20, 2000, the resolution of the Presidium of the High Attestation Commission of Ukraine № 13-08/9 included in the passport of the specialty “05.13.03 - systems and control processes” systems of intellectual decision support in conditions of uncertainty in the management of technological processes and complexes. This resolution also approved the passport of the specialty “05.13.23 - systems and means of artificial intelligence” which includes Specialty formula: Systems and means of artificial intelligence - a branch of science that deals with theoretical research, development and application of algorithmic and software-hardware systems and complexes with elements of artificial intelligence and modelling of human intellectual activity.

On December 14, 2006, by the Resolution of the Presidium of the High Attestation Commission of Ukraine № 31-06 / 11, the methods of artificial intelligence in economics were included in the passport of the specialty “08.00.11 - Mathematical Methods, Models and Information Technologies in Economics”.

According to the List of scientific specialties (order of the Ministry of Education and Science, Youth and Sports of Ukraine № 1057 of September 14, 2011) technical sciences were supplemented by the specialty “05.13.22 - systems and means of artificial intelligence”. legal artificial intelligence criminal

On August 30, 2017, by order of the Cabinet of Ministers of Ukraine № 600-r technology of artificial intelligence and robotics was included in the List of critical technologies in the field of armaments and military equipment.

On October 18, 2017, the Government of Ukraine (Cabinet of Ministers of Ukraine) adopted Resolution № 980 “Some issues of determining medium-term priority areas of innovation activity at the sectoral level for 2017-2021” in Section 8 of which provided as follows: development and implementation of artificial intelligence systems, including ¦ new intelligent transport technologies (unmanned vehicles, traffic management and planning in the city); ¦ technologies, algorithms and software and hardware of intelligent services for household, medical, social purposes; ¦ intelligent military systems (soldiers of the future, mobile demining works, intelligent weapons control systems); ¦ intelligent control systems for autonomous robots and robotic systems; ¦ intelligent decision support systems in conditions of uncertainty; ¦ pattern recognition systems (technical vision, speech, etc.); ¦ intelligent web technologies, cloud computing.

It should be noted that these provisions remain only on paper and have not yet been fully implemented as of the end of 2021.

The decision of the Council of the National Bank of Ukraine № 9-rd of March 31, 2020 recommended the launch of a program with testing of artificial intelligence methods in forecasting work with nonlinear processes.

The order of the Cabinet of Ministers of Ukraine № 1175-r of September 29, 2021 for the period 2022-2025 in the field of medicine plans to develop systems to support clinical solutions, personalized medicine, telemedicine, systems for big data processing (Big Data), artificial intelligence - engineering processing, use and acquisition of new knowledge using the model and data of the electronic health care system and related systems.

On December 2, 2020, the order of the Cabinet of Ministers of Ukraine № 1556-r approved the “Concept for the development of artificial intelligence in

Ukraine”. The terms are used in the following sense: artificial intelligence - an organized set of information technologies, using which it is possible to perform complex tasks by using a system of scientific research methods and algorithms for processing information obtained or independently created during work, as well as create and use their own knowledge bases, decision-making models, algorithms with information and identify ways to achieve the objectives; branch of artificial intelligence - the direction of activity in the field of information technologies which provides creation, introduction and use of technologies of artificial intelligence.

2. Is Ukraine using AI-based systems for predictive policing? Order of the Ministry of Justice of Ukraine № 3184/5 of 15.09.2020 regulates the use of the automated system “Cassandra” to assess the risks of criminal behaviour of convicts. This document amends other regulations, in particular, ¦ Instruction on the work of departments (groups, sectors, senior inspectors) of control over the execution of court decisions of penitentiary institutions and pre-trial detention centers (order of the Ministry of Justice of Ukraine of June 8, 2012 № 847/5), ¦ Procedure for compiling a pre-trial report (order of the Ministry of Justice of Ukraine of January 27, 2017 № 200/5), ¦ Procedure for the formation and maintenance of the Unified Register of Convicts and Detainees (Order of the Ministry of Justice of Ukraine of June 26, 2018 № 2023/5).

The Cassandra system on the basis of automated inference algorithms provides the formation of risk assessment of re-offending. The data required for risk assessment are entered by the registrar. Determining the degree of risk of recidivism (low, medium, high, very high) is carried out by the subsystem Cassandra automatically on the basis of the algorithm for calculating points and compliance with the amount of points to the established degrees of risk. The Cassandra provides a forecast of a person's re-offending, which is based on the use of machine learning and algorithms for automated conclusions (forecasts), based on the results of processing large structured data sets.

The forecast of a person's re-offending is formed for the purpose of research, comparing the results of the forecast with the result of risk assessment and improving the risk assessment system.

The Department of Artificial Intelligence of Kharkiv National University of Radio Electronics together with the Main Directorate of the National Police in Kharkiv region within the framework of the agreement on scientific and technical cooperation launched a project in the field of data mining based on modern neuro- fuzzy computational intelligence technologies.

The research results are implemented in the information-analytical complex “RICAS - Real-time intelligence crime analytics system” (http://ricas.org). It is a unique intelligent data analysis system that combines the basic and most modern methods and techniques of criminal analysis and real-time analytical search, which can significantly increase the efficiency and effectiveness of crime detection in hot pursuit, as well as previously unsolved crimes. RICAS allows you to perform the following types of analysis: * crime pattern analysis, * general profile analysis,

• analysis of a specific investigation (case analysis), * comparative analysis,

• offender group analysis, * specific profile analysis, * investigation analysis.

3. Is there a different legal definition of Machine Learning in Ukrainian law? There is no legal definition of Machine Learning in the national legislation and case law (decisions of the European Court of Human Rights and the Constitutional Court of Ukraine). Meanwhile, in the information space there is a training course “Machine Learning” by Prometheus¹. This course provides a broad view of the field of machine learning. Main topics of the course: The problem of learning. Training and testing. Generalization theory. Characteristic descriptions and types of quality functionality. Decision trees. Linear regression. Logistic regression. Support Vector Machines. Clustering and dimensionality reduction. Introduction to neural networks. Learning without a teacher. Reinforced training. Modern libraries of machine learning.

The Machine Learning was developed within the framework of the Initiative for the Development of think tanks in Ukraine, implemented by the Renaissance Foundation in partnership with the Foundation for the Development of Analytical Centres (TTF) with the financial support of the Swedish Embassy in Ukraine. The views and opinions expressed in this course are those of the author and do not necessarily reflect the position of the Government of Sweden.

4. Does Ukrainian law confer legal personhood or legal capacity to the AI systems? The national legislation of Ukraine currently does not confer legal personhood or legal capacity to the AI systems. In the process of working on the new Criminal Code of Ukraine, the authors of the report were part of an advisory group to prepare a section on crimes against information security. The position of the advisory group was discussed at the International Scientific Conference “Special Part of the Criminal Code of Ukraine: System and Content”, October 2022, 2021.

According to M. V. Karchevskyi Machine Learning / Prometheus, IRF: ML 101. URL: https://courses.prometheus.org.ua/

courses/IRF/ML101/2016_T3/about. Karchevskyi, Mykola (2012) “Kriminalno-pravova okhorona informaziynoi bezpeki Ukraini” [Criminal law protection of information security of Ukraine] 2012 monografia / M V. Karchevskyi; MVS Ukraini, Luganskiy dergavniy universitet im. E. O, Didorenka [Ministry of Internal Affairs of Ukraine, Luhansk State University of Internal Affairs named after E.O. Didorenko]. Lugansk: RVV LDUVS im. E. O. Didorenko, 2012. p. 263 (in Ukrainian); Karchevskyi, Mykola (2016) “Perspektivnie zadachi ugolovnogo prava v kontekste razvitia robototekhniki” [Perspective tasks of criminal law in the context of the development of robotics] 2016 Sozialna funksia kriminalnogo prava: problemi naukovogo zabezpechennia, zakonotvorennia ta pravozastosuvannia: materiali mignarodnoi naukovo-praktichnoi konferensii 12-13 govtnya 2016 roku / redkol.: V. Y. Taziy, V. I. Borisov ta insh. Kh.: Pravo, 2016. [The social function of criminal law: problems of scientific support, lawmaking and law enforcement: materials of the international scientific and practical conference, October 12-13. 2016]. p.p. 109-113 (in Ukrainian)., a specific requirement for the provisions of criminal law on crimes related to the use of new technologies should, not surprisingly, must be the technological neutrality. It is able to ensure the necessary stability of legislation in modern conditions of constant changes in technology. A striking example here is the current version of Part 3 of Article 190 of the

Criminal Code of Ukraine (fraud with the use of computer technology). How fair is it to consider fraud committed through illegal operations with the use of electronic computers as a serious criminal offense (sanction of Part 3 of Article 190 of the Criminal Code of Ukraine provides for punishment up to 8 (!) years in prison)? At the time of the inclusion of this norm in the Criminal Code of Ukraine (20 years ago), the use of computer technology to commit fraud could indeed indicate an increased public danger of encroachment. The prevalence of e-commerce and remote banking systems was insignificant. They have been used by large businesses. Therefore, the provisions of Part 3 of Art. 190 of the Criminal Code of Ukraine quite clearly outlined the range of acts that could reasonably be considered as a particularly qualified type of fraud, close in degree of public danger to fraud on a large scale. However, the rapid pace of penetration of information technology in the financial sector has led to a qualitative change in this type of fraud. Law enforcement agencies record a significant number of such crimes related to the infliction of harm, which corresponds to the signs of simple or qualified fraud (Part 1, Part 2 of Article 190 of the Criminal Code of Ukraine). Whether it is possible to consider reasonable, namely the interpretation of the norm, criminal-legal assessment of such actions under Part 3 of Art. 190 of the Criminal Code of Ukraine? The question is rather rhetorical. In modern conditions, there is no reason to say that the use of electronic computers in the process of fraud so increases the level of public danger of the act.

This example clearly demonstrates the thesis that in the conditions of rapid expansion of the scope of new technologies, the provisions of the Criminal Code are "tied" to certain technological aspects, will quickly lose relevance. They will acquire the character of those that do not reflect the objective level of social relations. That is why the inclusion in the Criminal Code of such concepts as “artificial intelligence”, “neurointerface”, “nanotechnology”, “genetic engineering” etc., should be considered critically. The highest legislative body of Ukraine (Verkhovna Rada of Ukraine) currently has no such registered bills.

But O. E. Radutniy Radutniy, Oleksandr (2017) “Criminal liability of the Artificial Intelligence” [2017] Problems of legality. 2017. Issue 138. http://plaw.nlu.edu.ua/artide/view/105661/106117; Radutniy, Oleksandr (2017) “Artificial Intelligence (shtuchniy intelekt) yak subektbpravovidnosin v galuzi kriminalnogo prava [Artificial Intelligence as a Subject of Legal Relations in the Field of Criminal Law].” U Politika v sferi borotbi zi zlochinnistu: Mizhnarodna naukovo-praktychna konferentsiia [International Scientific and Practical Conference “Policy on the Fight against Crime”]. Ivano-Frankivsk, 2017 (in Ukrainian); Radutniy, Oleksandr (2017) “Kriminalna vidpovidalnist shtuchnogo intelektu [Criminal Liability of Artificial Intelligence].” Informaziya i pravo 2 (2017): 124-33 (in Ukrainian). does not rule out the possibility of recognizing superartificial intelligence as a subject (persona) of legal relations, and therefore a subject (persona) of criminal offense. This prediction is based on important specific properties of super artificial intelligence: 1) the ability to think abstractly;

2) perception and recognition of all signals of the outside world; 3) obtaining most of the knowledge through training, as opposed to downloading the original data;

4) strategic thinking; 5) ability to deduction and induction, analysis and synthesis;

6) the ability to model the course of thoughts of the opponent; 7) the ability to work effectively in conditions of uncertainty and probability; 8) use of available information in the most expedient and optimal way, etc. But the main and defining features are the following: 1) awareness of the principles of their work and thus the ability to self-improvement (the first version forms an improved version of itself and so rewrites the program to infinity); 2) self-copying (ability to spread and selfpreservation); 3) solving the problem by brainstorming with the involvement of many copies of himself; 4) the ability to make decisions and act independently of human, and so on. However, the recognition of artificial intelligence as a subject (persona) of crime and legal relations will be expedient and justified only if the entire legal system is reformatted, including criminal law. Therefore, along with Section XIV-1 “Measures of a criminal nature against legal entities” of the Criminal Code of Ukraine, the possibility of a conditional section XIV-2 under the conditional title “Measures of a criminal nature against electronic persona (identity)” is not excluded (Radutniy, 2018) Radutniy, Oleksandr (2018) “Dodatkovi argumenti shodo pravosubeknosti shtuchnogo intelektu [Additional arguments on the legal personality of artificial intelligence].” In Internet rechey: problem pravovogo reguluvannya ta vprovadgennya: Mizhnarodna naukovo-praktychna konferentsiia [Internet of Things: Problems of Legal Regulation and Implementation], 46-50. Kyiv: Politekhnika, 2018 (in Ukrainian); Radutniy, Oleksandr (2018) “Mistze shtuchnogo intelektu v strukturi suspilnih vidnosin yaki ohoronyautsa kriminalnim pravom [The location of Artificial Intelligence in the Structure of Public Relations, which is Protected by Criminal Law].” In Fundamentalni problem kriminalnoi vidpovidalnosti: Naukoviy Polilog [Fundamental Problems of Criminal Liability: Scientific Polilog]. Kharkiv: Pravo 2018 (in Ukrainian)..

2. Existing criminal offences and criminalization

1. Have traditional offences and/or cybercrimes already been applied to illegal act committed by, through or against an AI system? The Unified State Register of Judgments of Ukraine (https://reyestr.court.gov.ua/) contains a small number of judgments that mention artificial intelligence, in particular:

• court decision of the Uzhhorod City District Court of the Zakarpattia Region of July 20, 2021 in case № 308/7867/21 - the level of danger to society of a person who has committed a criminal offense is determined by the automated system with artificial intelligence “Cassandra”;

• court decision of the District Administrative Court of Kyiv in the case № 640/20662/18 of July 1, 2020-on the error of the tax authority in calculating the debt to the taxpayer, which occurred using a computer program with partial participation of artificial intelligence;

• court decision of the Leninsky District Court of Kharkiv of August 25, 2021 in the case № 642/6961/19 - the use by the media of a news aggregator to cover the information picture of the day;

• court decision of the Commercial Court of Kharkiv region of May 28, 2020 in the case № 922/4148/19 - Innovative artificial intelligence company “Roofstreet”, for which the team Code&Care developed their product;

• court decision of the investigating judge of the Pechersk District Court of Kyiv of April 21, 2021 in case № 757/21307/21-k - pre-trial investigation established that a citizen of the People's Republic of China Person_1, who is the director of Golden Eg Technology, since the beginning of 2018 began to show considerable interest in projects, including those with limited access, in the field of nanotechnology, artificial intelligence, medical engineering, heat;

* court decision of the Kyiv Court of Appeal of August 18, 2021 in the case № 757/37395/21-k - the equipment seized during the search is used by MMI Engineering LLC within the main activities of the company, namely: AI (training of artificial intelligence; computer programming), maintenance of computer networks and other activities. This equipment has nothing to do with cryptocurrency mining.

In these court decisions, artificial intelligence acts as * the subject of the criminal offense (case № № 757/21307/21-k on information with limited access to certain technologies); * innovative product (cases № 922/4148/19 and № 757/37395/21-k); * aggregator of news and analytical content (case № 642/6961/19); * a computer program in the service of the state - the tax authority (case № 640/20662/18) or the probation authority (case № 308/7867/21).

In addition, in the judicial practice of Ukraine (see https://reyestr.court.gov.ua/Review/97444075 or -/87299694) there are cases when certain algorithmic bots are used in criminal offenses in the field of trafficking in narcotic drugs, psychotropic substances, their analogues or precursors. In particular, this applies to cases of creating a bot in the Telegram channel for the automatic sale of psychotropic substances and drugs on the Internet. In these cases, criminal offenses were committed with the help of an appropriate computer program with elements of artificial intelligence. Thus, the Telegram chatbot, according to pre-settings, automatically generated information for the buyer about the place, method of payment, type and amount of drugs, psychotropic substances that the buyer wanted to get (case № 686/8319/20 from the Common Register of Judgments of Ukraine).

However, there is currently an active fight against such chatbots, in particular, due to the alternative chatbot “StopNarcotics” in the system of the National Police of Ukraine. With its help 1583 addresses were blocked. Police have also launched a special bot called “DrugHunters”, which can be used to anonymously report about drug or drug dealers.

2. Has Ukrainian law introduced new offences related to designing, programming, developing, producing, functioning or making use of AI systems? The national legislation of Ukraine does not have any new corpus delicti that would explicitly mention AI. The so-called “computer crimes” (in the narrow sense) are provided for in Section XVI of the Criminal Code of Ukraine (Articles 361-363-1). The main statistical indicators of this group of crimes for the last 5 years (2016-2020) are the following: the number of recorded criminal proceedings (registered crimes) increased from 865 (2016) to 2498 (2020); the number of convicts increased from 24 (2016) to 56 (2020). The vast majority of persons were convicted of unauthorized interference with computer equipment (Article 361 of the Criminal Code of Ukraine) and illegal actions of persons entitled to access information (Article 362 of the Criminal Code of Ukraine). Starting from 2018, the number of convicted persons for selling or distributing malicious software or hardware (Article 361-1 of the Criminal Code of Ukraine) and information with limited access (Article 361-2 of the Criminal Code of Ukraine) is growing.

The national legislation of Ukraine does not have any new corpus delicti that would explicitly mention AI.

To date, there is no normative basis for recognizing AI as a subject of crime. According to the first part of Article 18 of the Criminal Code of Ukraine, the subject of a criminal offense is only a human sanity person who has committed a criminal offense at an age from which, in accordance with this Code, criminal liability may arise.

According to national legal doctrine, the AI system can be considered the object of a crime when it is part of the protected social relations, any AI system refers to methods or means of committing a crime.

3. Highlight whether they are crimes of mere conduct, commission and omission offences, consummate offence, crimes with intent, etc. According to Article 13 of the Criminal Code of Ukraine, a completed criminal offense is an act that contains all the elements of a criminal offense under the relevant article of the Special Part of this Code. An unfinished criminal offense is preparation for a criminal offense and attempt to commit a criminal offense.

According to Article 14 of the Criminal Code of Ukraine, preparation for a criminal offense is the search for or adaptation of means or tools, search for accomplices or conspiracy to commit a criminal offense, removal of obstacles, and other intentional creation of conditions for committing a criminal offense. Preparation for a criminal offense or a crime for which the article of the Special Part of this Code provides for imprisonment for up to two years or other, less severe punishment, does not entail criminal liability.

According to Article 15 of the Criminal Code of Ukraine, an attempt to commit a criminal offense is the commission by a person of direct intent of an act (action or omission) directly aimed at committing a criminal offense under the relevant article of the Special Part of this Code, if the criminal offense reasons that did not depend on her will. An attempt to commit a criminal offense is complete if the person has taken all the actions he or she deems necessary to complete the criminal offense, but the criminal offense has not been completed for reasons beyond his or her control. An attempt to commit a criminal offense is incomplete if the person, for reasons beyond his control, has not taken all the actions he considered necessary to bring the criminal offense to an end.

According to Article 24 of the Criminal Code of Ukraine, intent is divided into direct and indirect. Direct intent is if a person was aware of the socially dangerous nature of his action (action or inaction), foresaw its socially dangerous consequences and wanted them to occur. Indirect intent is if a person was aware of the socially dangerous nature of his action (action or inaction), foresaw its socially dangerous consequences and although he did not want to, but consciously assumed their occurrence.

According to Article 25 of the Criminal Code of Ukraine, negligence is divided into criminal illegal self-confidence and criminal illegal negligence. Carelessness is a criminal illegal self-confidence if a person foresaw the possibility of socially dangerous consequences of his action (action or inaction), but recklessly hoped to prevent them. Carelessness is a criminal unlawful negligence if a person did not foresee the possibility of socially dangerous consequences of his action (action or inaction), although he should and could have foreseen them.

4. Specify who can be considered the possible perpetrator and / or victim of the new AI offences (e.g. producers / programmers / system engineers / developers / designers etc.). As long as the AI is not endowed with self-improvement properties and does not make decisions autonomously from humans, it is possible to consider the developer or user as the perpetrator of the crime. The victims of these crimes are not fundamentally different from the victims of traditional crimes.

5. Indicate whether individual criminal liability requires a specific mental element and whether it involves also recklessness and / or negligence.

According to Article 24 of the Criminal Code of Ukraine, intent is divided into direct and indirect. Direct intent is if a person was aware of the socially dangerous nature of his action (action or inaction), foresaw its socially dangerous consequences and wanted them to occur. Indirect intent is if a person was aware of the socially dangerous nature of his action (action or inaction), foresaw its socially dangerous consequences and although he did not want to, but consciously assumed their occurrence.

According to Article 25 of the Criminal Code of Ukraine, negligence is divided into criminal illegal self-confidence and criminal illegal negligence. Carelessness is a criminal illegal self-confidence if a person foresaw the possibility of socially dangerous consequences of his action (action or inaction), but recklessly hoped to prevent them. Carelessness is a criminal unlawful negligence if a person did not foresee the possibility of socially dangerous consequences of his action (action or inaction), although he should and could have foreseen them.

6. Could the legal persons be held liable for AI crimes committed by any person acting individually or having a leading position within the legal person? According to the Criminal Code of Ukraine, a legal entity is not a subject (persona) of a crime, accordingly, it cannot bear legal responsibility. But with regard to legal entities, the Criminal Code of Ukraine contains Section XIV-1 “Measures of a criminal nature against legal entities”.

7. Indicate whether there is any defence excluding the criminal responsibility of the perpetrator or of the legal person in order to avoid the risk of over- criminalization if the AI systems are produced, used or put on the market for legal purposes, e.g. for scientific or research reason. Today, the liability of the executor or legal entity in AI matters is resolved in a general way on the basis of traditional norms. In the process of working on the new Criminal Code of Ukraine, the authors of the report were part of an advisory group to prepare a section on crimes against information security. The position of the advisory group was discussed at the International Scientific Conference “Special Part of the Criminal Code of Ukraine: System and Content”, October 20-22, 2021.

According to M. V. Karchevskyi Karchevskyi, Mykola (2018) “Pravove reguluvannya sozializazii shtuchnogo intelektu” [Legal regulation of socialization of artificial intelligence] 2017. 2(78) Visnik Luganskogo dergavnogo universitety im. E. O. Didorenka [Bulletin of Luhansk State University of Internal Affairs named after E. O. Didorenko]. 99-108 (in Ukrainian)., a specific requirement for the provisions of criminal law on crimes related to the use of new technologies should, not surprisingly, must be the technological neutrality. It is able to ensure the necessary stability of legislation in modern conditions of constant changes in technology.

8. Whether reports or legal literature suggest the introduction of new criminal offences linked to AI systems (please provide also bibliographic references). Today, it is clear that strong artificial intelligence is a hypothetical technology and will remain at this stage indefinitely. It is possible that technologies of strong artificial intelligence will take the status of a subject (persona) of law, then new spheres of justice will appear. In addition to traditional justice, we can talk about the emergence of two new types, conditionally call them “mixed justice” and “artificial intelligence justice” (M. V. Karchevskyi, 2019) Karchevskyi, Mykola (2019) “Perspektivi pravovogo reguluvannya v konteksti gipotezi rozvitku tekhnologiy transgumanizmu” [Perspectives of legal regulation in the context of the hypothesis of the development of transhumanism technologies] 2019. 1(87) Visnik Luganskogo dergavnogo universitety im. E O Didorenka [Bulletin of Luhansk State University of Internal Affairs named after E.O. Didorenko]. 115-127 (in Ukrainian).. Mixed justice will include forms of resolving legal disputes between individuals, legal entities, society and businesses. The justice of artificial intelligence will include forms of resolving legal disputes between robots (or unit of artificial intelligence). In addition, the functioning of this system of justice will counteract the work that threatens social development and stability. Most likely, the human justice system will not be copied for artificial intelligence. Fundamentally different physical characteristics and needs require a priori to abandon this approach. At the same time, the creation of this system will be a necessary condition to provide humanity with the opportunity to control the development of social processes. Most likely, the justice of artificial intelligence will be created on the basis of robots. The physical and intellectual data of a human, obviously, will be insufficient for the effective functioning of this system of justice. The creation of such a system will require generalization into clear algorithms of experience gained during the existence of traditional justice. Such a generalization may become one of the main directions of future legal science.

It is likely that changes in the judiciary will take place in a different scenario. For now, it is obvious only that the impact of strong artificial intelligence technologies on the fight against crime can be studied only hypothetically.

At the same time, it makes sense today to consider the issue of compliance of the current criminal legislation with the level of artificial intelligence technologies only in the context of “weak” artificial intelligence, because it actually exists.

Taking into account the clarifications made, the answer to the question is “yes”. The current Criminal Code of Ukraine is able to provide an adequate response to the consequences of the use of systems of “weak” artificial intelligence, is also able to provide adequate protection of rights and interests in the use of these technologies. To the draft of the new Criminal Code, O.E. Radutniy Radutniy, Oleksandr (2018) “Subyektnist shtuchnogo intelektu u kriminalnomu pravi [The Subjectivity of Artificial Intelligence in Criminal Law].” Pravo Ukraini 1 (2018): 123-36 (in Ukrainian); Radutniy, Oleksandr (2019) “Adaptation of criminal and civil law in view of scientific-technical progress (Artificial Intelligence, DAO and Digital Human Being)” [2019] Problems of Legality, No 144 (2019), [S.l.], n. 144, p. 138-152, March 2019. http://plaw.nlu.edu.ua/article/view/155819/159365; Radutniy, Oleksandr (2019) “Morality and Law for Artificial Intelligence” [2019] Proceedings of the 1st International Symposium on Intellectual Economics, Management and Education, September 20, 2019. Vilnius Gediminas Technical University. Vilnius: Vilnius Gediminas Technical University, 2019. p. 44-46. proposed an article about illegal actions with the digital image of human persona.

9. Does your domestic law provide for positive obligations for persons and/or legal person designing, developing, producing, testing, selling or distributing AI systems? In general terms, such positive obligations are provided by the Order of the Cabinet of Ministers of Ukraine № 1556-r of December 2, 2020 “Concept of development of artificial intelligence in Ukraine”.

The main task of state policy in the field of legal regulation of artificial intelligence is to protect the rights and freedoms of participants in relations in the field of artificial intelligence, development and use of artificial intelligence technologies in compliance with ethical standards.

In order to achieve the goal of the Concept in this area, the following tasks should be ensured: ¦ implementation of the norms enshrined in the “Recommendations on Artificial Intelligence” adopted in June 2019 by the Organization for Economic Cooperation and Development (OECD / LEGAL / 0449), subject to the ethical standards set out in Recommendations CM / Rec (2020) 1, approved 8 April 2020 by the Committee of Ministers of the Council of Europe for member states on the impact of algorithmic systems on human rights in the legislation of Ukraine; ¦ elaboration of the issue of compliance of the legislation of Ukraine with the guiding principles established by the Council of Europe on the development and use of artificial intelligence technologies and its harmonization with the European one; ¦ assessment of the possibility and determination of the limits (ethical, legal) of the use of artificial intelligence systems for the purposes of providing professional legal assistance; ¦ ensuring the functioning and operation of technical committees of standardization in accordance with the requirements of 7.1.5 DSTU 1.14: 2015 “National standardization. Procedures for the establishment, operation and termination of technical committees of standardization "in the field of artificial intelligence; ¦ ensuring cooperation between the relevant Technical Committees of Ukraine and the international subcommittees of standardization ISO / IEC JTC 1 / SC 42 Artificial Intelligence on the joint development of standards in the field of artificial intelligence; ¦ support for initiatives to create organizational forms of cooperation between interested legal entities and individuals in the field of artificial intelligence; ¦ development of a Code of Ethics for artificial intelligence with the participation of a wide range of stakeholders.

3. Applicability of Traditional Criminal Law Categories

1. According to Ukrainian law and / or jurisprudence, is the AI system considered as a “computer system” as defined by Article 1, lett. of Cybercrime Convention and / or Article 2, lett. a) of Directive EU/2013/40? On December 2, 2020, the order of the Cabinet of Ministers of Ukraine № 1556-r approved the “Concept for the development of artificial intelligence in Ukraine”. In this Concept, the terms are used in the following sense: artificial intelligence - an organized set of information technologies, using which it is possible to perform complex tasks by using a system of scientific research methods and algorithms for processing information obtained or independently created during work, as well as create and use their own knowledge bases, decision-making models, algorithms with information and identify ways to achieve the objectives; branch of artificial intelligence - the direction of activity in the field of information technologies which provides creation, introduction and use of technologies of artificial intelligence.

2. Are there specific problems with respect to the principle of legality?

Such problems do exist. However, they are not related to AI. They are caused in most cases by a distorted practice of applying the law, which at the level of the legal norm is formulated quite progressively and constructively in most cases. According to the Criminal Code of Ukraine “No one can be held responsible for acts that at the time of the commission were not recognized by law as an offense” (Part 2 of Article 58). The content of this principle is also disclosed in Part 2 of Art. 4 of the Criminal Code of Ukraine, according to which “the criminality and punishment of an act are determined by the law on criminal liability, which was in force at the time of the act”. This provision, firstly, obliges public authorities and officials to strictly adhere to the rules of criminal law in prosecuting and punishing a person and, secondly, excludes the application of criminal law by analogy with actions that are not provided for norms of the General and Special Parts of the Criminal Code of Ukraine.

3. Is analogy admissible? Has it been used in order to criminalize illegal acts related to AI systems? The application of the law on criminal liability by analogy is prohibited by Part 4 of Art. 3 of the Criminal Code of Ukraine. A classic example is the prohibition of analogy in criminal law, as a result of which the composition of crimes cannot be established solely by judgment by analogy or other supplement to the rule of law (the principle of “nulla poena sine lege”).

4. Are the provisions concerning attempted crime applicable to AI- related crimes? Are there already cases of AI-related crimes qualified as attempted crimes? According to the Criminal Code of Ukraine, the provision on attempted crime is applied to the subject (persona) of the crime in case of committing an act directly aimed at committing a crime, failure to complete the crime, the reasons for incompleteness of the crime do not depend on the will of the perpetrator. Today, artificial intelligence is not considered a subject (persona) of crime. Therefore, in all the above cases, artificial intelligence can be used as a tool or method of committing a crime.

5. Is it possible to apply existent rulings of joint-perpetration and participation in the commission of the crime to AI related crimes? Who can be considered a joint-perpetrator or participant in the commission of the crime (please refer to both human and artificial agents)? Is the “perpetration- by-another” liability model applicable? According to the Criminal Code of Ukraine, complicity in a criminal offense is the intentional joint participation of several subjects of a criminal offense in the commission of an intentional criminal offense. AI is not the subject (persona) of a crime, so it cannot be an accomplice, perpetrator, organizer, instigator or accomplice.