The Role of Artificial Intelligence in Criminal Justice - Reality and Perspective

ანოტაცია

Artificial intelligence has recently become one of the hot topics in scientific and professional circles of various fields. Current technological processes in the world have significantly increased the use of digital technologies in practically all fields. Over the past century, several researchers and mathematicians have been developing the idea that computing machines can not only perform typical technical tasks but also learn to think and perform individual tasks accordingly, like humans. This idea has developed over time, and nowadays, the main topic among scientists, researchers, and practitioners worldwide is artificial intelligence and the tasks it can perform. Artificial intelligence is already actively used in various fields of public activity and many professions, and there is still talk that many professions can be replaced by artificial intelligence in the future. In this regard, using artificial intelligence in jurisprudence is particularly important and controversial. Based on the relevance of the mentioned issue, this article is dedicated to the use of artificial intelligence in the field of criminal proceedings and investigations. The article discusses the origin and development of artificial intelligence and its primary role in criminal proceedings and investigations. In the work, special attention is paid to the experience of different countries. The paper also analyzes court decisions of precedent importance in this regard. The paper also contains the main challenges of using artificial intelligence in criminal justice and investigation.

Keywords: Justice, ethics, technology.

Introduction

Nowadays, artificial intelligence and its role in various fields is one of the most relevant issues, and advice is given in almost all directions. With the development of technologies, it becomes increasingly important to set complex and large-scale tasks for artificial intelligence.

The issue has become particularly relevant in recent years. Various technologies based on artificial intelligence are improving more and more. In scientific and professional circles, the opinion is often heard that in a fairly short period, people will be replaced by the so-called Thinking Machines, and many scientists even express the opinion that many professions will no longer exist. Obviously, this will not be the first time for humanity; the history of human existence testifies that many professions have been replaced by technological progress, which is not strange. However, replacing a judge with a thinking machine is still strange. Many works can be found on the use of artificial intelligence in law. Still, this time, based on the issue’s relevance, our goal is to study in what part artificial intelligence can be used in criminal proceedings and investigations. To achieve this goal, the essence of artificial intelligence will be studied within the framework of the research, its capabilities, the area of ​​its use, the existing practice of use, the approaches of different countries will be analyzed, and the author’s position on the following issues will be presented at the end of the paper:

• Main challenges and weaknesses of using artificial intelligence in criminal litigation and investigation;


• The current area of ​​use of artificial intelligence in criminal proceedings and investigation and the perspective of its further use. 
  
  1. The History of Creation and Development of Aritifical Inteligence

A few people are considered to be the founders of artificial intelligence itself. First of all, Alan Turing’s contribution is noteworthy in this regard. He is often referred to as the father of artificial intelligence. Turing’s 1950 scientific work “Computing Machine and Intelligence” is noteworthy in this regard. The paper presented the concept of a computing machine with human intelligence. Turing’s thinking laid the foundation for further development of artificial intelligence - i.e. To create a computing machine that thinks like a human.[1] Later, the topic was developed by John McCarthy, a professor of mathematics, who used the term “artificial intelligence” in 1956. It was under his organization that the Darthmuz conference was organized in 1956, where the phenomenon known to us today was called artificial intelligence, and this practically laid the foundation for the development of this great direction, thus artificial intelligence even became a separate discipline.[2] McCarthy co-founded the MIT AI LAB with Marvin Minks, and in the fields of robotics and cognitive psychology, Minks made virtually inestimable contributions to the development of artificial intelligence.[3] Later, Herbert Simon and Allen Newell developed early artificial intelligence programs that were focused on solving human problems. This laid the foundation for the “thinking” of artificial intelligence, according to how it would be possible to use human cognitive skills for a computing machine.[4] Arthur Samuel, who created one of the first self-learning programs in the fifties of the last century, proved that a computer can learn from experience.[5] The analysis of modernity reveals that in the development of artificial intelligence, Geoffrey Hinton, Joshua Bengio and Ian Lekun have also made special contributions, and that’s why Hinton is called the godfather of artificial intelligence.[6]

In addition to science and researchers, organizations have made the greatest contribution to the development of artificial intelligence. Thanks to an IBM project and artificial intelligence, in 1997, chess champion Garry Kasparov was defeated by artificial intelligence, proving that it can be used in strategy games.[7] Google DeepMindm did a similar project with Go World Champion Lee Sedol in 2016.[8]

2. The Role of Artificial Intelligence in Crime Prediction and Analysis

Along with the development of artificial intelligence, an irreversible process of its application in various fields has started step by step, and law is no exception in this regard. From an early stage of development, the field of criminal justice and investigations has used artificial intelligence as part of investigative data analytics to make it easier for investigative agencies to manage critical operational investigative information and identify suspects by following data patterns.[9] Early applications of artificial intelligence in criminal justice included the use of data analytics to manage and analyze crime databases. For example, the Federal Bureau of Investigation (FBI) created the National Crime Information Center (NCIC) in the 1960s, which used computer systems to efficiently store and retrieve criminal records.[10] Since 1990, artificial intelligence has been actively used for prediction in criminal proceedings and the fight against crime.[11] Law enforcement agencies have begun experimenting with predictive policing methods. These models used statistical techniques to predict future crime hotspots. This model laid the foundation for today’s sophisticated forecasting systems using artificial intelligence.[12] The IBM organization also made a special contribution in this case and, in cooperation with various police departments, implemented a number of important projects in the creation of a crime analysis and prediction system using artificial intelligence.[13]

3. Artificial Intelligence in Contemporary Criminal Proceedings

Nowadays, artificial intelligence is used even more actively in criminal proceedings and investigations. Artificial intelligence in criminal law involves a technological process (algorithms, language processing, computer vision) designed to perform tasks that normally require human intelligence, human intervention. All these technological processes are focused on increasing the efficiency of the investigative and justice process. In this regard, even today, one of the main missions of artificial intelligence in criminal law is prediction.[14] In particular, analysis of existing data based on various sources, for future crime prediction and prevention. A good example is the experience of the Los Angeles Police Department (LAPD), which introduced the Predpol program for more efficient use of police resources and, accordingly, to increase the efficiency of police work. The Chicago Police Department uses artificial intelligence to identify individuals who are at high risk of committing violence. This has helped to prevent violence across the state.[15]

Artificial intelligence systems can process large amounts of data better than humans, so it is often used to process such voluminous investigative information. For example, the New York Police Department (NYPD) uses pattern recognition software to analyze information important to investigations and identify serial criminals, which helps them work more efficiently.[16]

We must not forget the research part as well. Artificial intelligence is a very effective tool for research. Even LexisNexis or HeinOnline helped lawyers and judges find the necessary information in their time. It is safe to say that artificial intelligence has made it even easier. In some jurisdictions, the courts use artificial intelligence as a decision-making tool. This does not apply to all decisions. Such systems analyze past cases and outcomes to provide recommendations to reduce bias and improve the consistency of judicial decisions, ensuring uniformity of judicial practice. However, the use of artificial intelligence in this field has led to differences of opinion about the risks of algorithmic bias and the transparency of decision-making processes.^[17] No less interesting is the Estonian experience, which created a judge with artificial intelligence, which aims to consider small disputes, which ultimately significantly relieves the judicial system.[18]

4. The Analysis of Different Countries’ Experiences

Comparative studies prove that the tasks described above can be performed smarter and more effectively using artificial intelligence, which improves the quality of the work performed. Nevertheless, it should be noted that using artificial intelligence to perform such tasks poses a rather acute ethical issue. It is about the transparency and accountability of decisions based on artificial intelligence.[19] It is impossible not to mention here that one of the biggest challenges in processing important information for legal proceedings through artificial intelligence is the issue of personal data protection. Although the use of artificial intelligence in criminal proceedings and investigations is at first glance a novelty, many countries already have many years of experience in using artificial intelligence in this field. In this regard, first of all, the experience of the United States of America we described should be mentioned above. It has been using artificial intelligence for a long time in crime prediction and prevention. Among them, as mentioned above, special programs (Predpol, Compstat) are used for crime analysis and management.^[20]

The People’s Republic of China is a leading country in terms of the most widespread and unrestricted use of artificial intelligence. Artificial intelligence is used for total surveillance in China, one of the goals of which is crime prevention. Thousands of so-called facial recognition devices are installed throughout the country, and a smart camera is integrated with the social credit system, thus ultimately evaluating the citizens.

Obviously, this system of assessment goes beyond the goals of criminal law, and it is also undisputed that this system of surveillance has developed, especially in China, to the extent that prohibitions and restrictions in relation to democratic values ​​are not found in this country.[21]

As for European countries, in the wake of the development of artificial intelligence, more and more attention is being paid to personal data protection regulations in these countries in order to exclude the misuse of artificial intelligence that would violate human rights and affect people’s personal lives, etc. For example, United Kingdom police agencies use AI-based programs to predict crime, subject to ethical and legal constraints. Germany actively uses artificial intelligence in legal research, in full compliance with the principles of good faith. France, Japan, South Korea, and India also use AI for crime prediction/analysis and scientific research, subject to ethical and legal constraints.[22]

5. The Analysis of Judicial Practice

Obviously, in this case, a certain practice was adopted. In 2016, a court case was reviewed in the United States of America: Loomis v. Wisconsin. Eric Loomis was sentenced to 6 years in prison based on a risk assessment calculated using the Compas (Correctional Offender Management Profiling for Alternative Sanctions) algorithm. Loomis disputed using this program and algorithm to determine his fate, appealing that it violated his due process rights because the evaluation process was completely opaque to him, and it remained unclear how the algorithm arrived at such a risk score. The court did not share the person’s position based on the argumentation that this was not the only indicator on which the court made its decision. Therefore, although the appellant’s request was not satisfied, the case is precedent based on this argumentation.[23] No less precedential was the case from the UK judicial practice - R. (Bridges) v. Chief Constable of South Wales police. In 2020, a person challenged South Wales Police’s use of a facial recognition system, claiming it breached his right to privacy, the European Convention on Human Rights and the Data Protection Act. The court considered the appellant’s opinion and granted his request, noting that the police did not realize how large and harsh the impact of such systems was on people’s privacy and confidentiality and the protection of personal data. This decision also sets a precedent in the sense that despite the great influence of technological processes and artificial intelligence in the world, it is necessary to maintain a balance to avoid the abuse of rights and gross infringement of personal data.[24] Speaking about practice, it is impossible not to analyze the practice of the European Court of Human Rights. In one of the decisions (the case related to the situation when the bank evaluated credit applications through a special program), the court noted that transparency and accountability in the decision-making process are important when using artificial intelligence. An individual has the right to have an automated decision reviewed by a human. Automatic decisions that significantly affect the fate of a person must be explained and justified.^[25]

Conclusion

Therefore, there is no doubt that the creation and development of artificial intelligence results from the joint work of many scientists and organizations, and many people and organizations have been involved in this chain. Alan Turing and John McCarthy are still considered pioneers in this field because they laid the foundation for such capabilities of the computing machine as the idea. Subsequently, Hinton’s team’s contribution to developing this idea is immeasurable, and finally, such large organizations as IBM and Google DeepMind have better demonstrated the capabilities of artificial intelligence in practice.

Therefore, it seems that a large number of countries are using artificial intelligence in criminal proceedings, investigations and scientific activities. However, their attitude is different. For example, China focuses more on population control and does not consider human rights; in the United States of America, it is mostly used in crime analysis and prediction. Obviously, the country pays a lot of attention to human rights. However, in this case, in terms of personal data, we do not have such strict regulations here as in many European countries, where artificial intelligence is also used, although the greatest attention is paid to human rights and personal data. In European countries, human rights and privacy are considered to be of paramount interest when using artificial intelligence technologies.

The cases we discussed above illustrate the growing trend of using artificial intelligence in criminal proceedings and investigations. In general, the justice systems of various countries support the use of artificial intelligence only on the condition that the process is transparent and that it is not the only prerequisite for determining substantive legal consequences. It is also important that artificial intelligence systems are free from bias and take into account the rights of individuals - the inviolability of their private lives.

Although technologies and programs based on artificial intelligence have greatly facilitated the tasks in criminal proceedings and investigations, it is clear that we face certain risks and challenges. First of all, in the conditions of such systems, the probability of error is quite high. Therefore, it is very important to assemble and generate them correctly. For the system’s proper functioning, it is very important to correctly generate and reflect the existing data because a mistake made here can lead to an incorrect or discriminatory decision. Unlike humans, artificial intelligence will not be self-critical about its own decisions. The biggest challenge is privacy and personal data; when using artificial intelligence systems on a massive scale, it is important to maintain the necessary balance and data security. One of the ethical issues is that the decisions made by this system should be transparent, and there should be a system of accountability. The use of artificial intelligence in criminal law in different jurisdictions is associated with substantial challenges, as legal standards, regulations, and cultural attitudes towards artificial intelligence vary widely. Along with the development and improvement of artificial intelligence in the field of criminal proceedings and investigation, it is expected:

• To improve the police forecasting systems, to create even more complete algorithms that will combine a large amount of data with the help of which the algorithm will be able to perform the closest to reality, perhaps the most inaccurate, forecast in this regard better than people and criminologists;


• Artificial intelligence technologies will be used more and more in the examination, so it will be possible to obtain and attach more evidence, and this evidence will be more accurate and reliable;


• Artificial intelligence technologies will increasingly make it possible to restore the most accurate crime scene or route, which will obviously make the response to it and the investigation process more complete;


• Artificial intelligence technologies will facilitate the management of inmates in prisons, including tracking their behavior and planning personalized responses, including rehabilitation.

The extent to which artificial intelligence can replace judges is a separate discussion. In addition to the ethical objections discussed above, another objection arises. In particular, we are talking about such a necessary component of implementing justice as the inner faith of the judge. Could a computing machine have internal beliefs? Of course it can, although it may have an internal belief such that humans again write it with an appropriate algorithm. Will the calculating machine be able to evaluate such circumstances as extreme cruelty? Sincere repentance of actions by a person, etc. The answer to this question is rather vague and ambiguous. It is clear that artificial intelligence technologies are improving more and more, and their application area is expanding more and more. It is also undisputed that the idea looks quite attractive in the background of an overburdened judicial system. However, the real perspective of its implementation is rather vague at this stage.

Bibliography

References/Monographs

1. Dormehl, L. (2017). Thinking Machines. USA, New York, TarcherPerigee;


2. Ashley, K. D. (2017). Artificial intelligence and legal analytics. Cambridge University press, UK;


3. Kelly, W. R. (2016). The future of crime and punishment: Smart policies for reducing Crime and saving money. Rowman & Littlefield Publishers, USA;


4. Antunes, H. S., Freitas, P. M., Oliviera, A. L., Pereira, C. M., Sequiera, E. V., Xavier, L. B. (2023). Multidisciplinary Perspectives on Artificial intelligence and the Law. Springer, Spain;

Scientific Articles

1. Turing, A. M. (1950). Computing machinery and intelligence. Mind, Volume LIX, Issue 236;


2. Moor, J. (2006). The Dartmouth College Artificial Intelligence Conference: The Next Fifty Years. AI Magazine, Volume 27;


3. Gugerty, L. (2006). Newell and Simon’s Logic Theorist: Historical Background and Impact on Cognitive Modeling. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 50(9);


4. Samuel, A. (1992). Arthur Samuel: Pioneer in machine learning. BM Journal of Research and Development, 36(3);


5. Newborn, M. (2000). Deep Blue’s contribution to AI. Annals of Mathematics and Artificial Intelligence, 28(1-4);


6. Gund, P., Phalke, V. S. (2023). Investigating crime a role of Artificial intelligence in criminal Justice. The Online Journal of Distance Education and e-Learning, Vol. 1;


7. Kabol, A. F. (2022). The Use Of Artificial Intelligence In The Criminal Justice System (A Comparative Study). Webology, 19(5);


8. Bendouzid, B. (2019). To predict and to manage. Predictive policing in the United States. Big Data & Society, 6(1);


9. Vasdani, T. (2023). From Estonian AI Judges to robot mediators in Canada. U.K, The Lawyer’s Daily, Vol. II;


10. Wischmeyer, T. (2020). Artificial intelligence and transparency: Opening the black box. Regulating Artificial Intelligence;


11. Mugari, I., Obioga, E. (2021). Predictive Policing and Crime Control in The United States of America and Europe: Trends in a Decade of Research and the Future of Predictive Policing. Social Sciences, 10(6);


12. State v. Loomis. (2016). Harward law review. Vol. 130, Issue 5, Harward University;


13. Szappanyos, M. (2023). Artificial intelligence: Is the European court of Human rights prepared? Acta Humana, Vol. 11(1).

Webpages

1. MIT Management Sloan School. (23.03.2023). Why neural net pioneer Geoffrey Hinton is sounding the alarm on AI <https://mitsloan.mit.edu/ideas-made-to-matter/why-neural-net-pioneer-geoffrey-hinton-sounding-alarm-ai> (Last access: 07.2024).


2. Google DeepMind. (02.04.2023). Alphago <https://deepmind.google/technologies/alphago/> (Last access: 07.2024).


3. Banafa A. (2023). The use of AI in the criminal justice field https://www.linkedin.com/pulse/use-ai-criminal-justice-field-prof-ahmed-banafa-mrtnf (Last access: 20.07.2024).


4. Carpertner, W. (2023). Presidents’ massage: AI’s transformative impact <https://www.policechiefmagazine.org/presidents-message-ais-transformative-impact/> (Last access: 07.2024).


5. Yucel, G. (2024). Crime prevention effort with artificial intelligence in China: analysis <https://www.linkedin.com/pulse/crime-prevention-efforts-artificial-intelligence-china-gokcen-yucel-aascf> (Last access: 06.2024).


6. Industry views. (2024). Mahalias, I. AI adoption in criminal justice – How can industry support the justice system in implementing Artificial intelligence <https://www.techuk.org/resource/ai-adoption-in-criminal-justice-how-can-industry-support-the-justice-system-in-implementing-artificial-intelligence.html> (Last access: 07.2024).


7. Hunton, A. K. (2020). UK court of appeal finds automated facial recognition technology unlawful in Bridges v South Wales police <https://www.huntonak.com/privacy-and-information-security-law/uk-court-of-appeal-finds-automated-facial-recognition-technology-unlawful-in-bridges-v-south-wales-police> (Last access: 07.2024).


8. Rigano, G. (2018) Using artificial intelligence to address criminal justice needs. National institue of justice (NIJ) https://nij.ojp.gov/topics/articles/using-artificial-intelligence-address-criminal-justice-needs (Last access: 20.07.2024).

Footnotes

^[1] Turing, A. M. (1950). Computing machinery and intelligence. Mind, Volume LIX, Issue 236, pp. 433-460.

^[2] Moor, J. (2006). The Dartmouth College Artificial Intelligence Conference: The Next Fifty Years. AI Magazine, Volume 27, pp. 87-91.

^[3] Dormehl, L. (2017). Thinking Machines. USA, New York, TarcherPerigee, pp. 9-10.

^[4] Gugerty, L. (2006). Newell and Simon’s Logic Theorist: Historical Background and Impact on Cognitive Modeling. Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 50(9), pp. 880-884.

^[5] Samuel, A. (1992). Arthur Samuel: Pioneer in machine learning. BM Journal of Research and Development, 36(3), pp. 329-33.

^[6] MIT Management Sloan School. (23.03.2023). Why neural net pioneer Geoffrey Hinton is sounding the alarm on AI <https://mitsloan.mit.edu/ideas-made-to-matter/why-neural-net-pioneer-geoffrey-hinton-sounding-alarm-ai> (Last access: 20.07.2024).

^[7] Newborn, M. (2000). Deep Blue’s contribution to AI. Annals of Mathematics and Artificial Intelligence, 28(1-4), pp. 27-30.

^[8] Google DeepMind. (02.04.2023). Alphago <https://deepmind.google/technologies/alphago/> (Last access: 25.07.2024).

^[9] Gund, P., Phalke, V. S. (2023). Investigating crime a role of Artificial intelligence in criminal Justice. The Online Journal of Distance Education and e-Learning, Vol. 1. pp.50-56.

^[10] Kabol, A. F. (2022). The Use Of Artificial Intelligence In The Criminal Justice System (A Comparative Study). Webology, 19(5), p. 593.

^[11] Ashley, K. D. (2017). Artificial intelligence and legal analytics. Cambridge University press, UK, pp. 60-62.

^[12] Bendouzid, B. (2019). To predict and to manage. Predictive policing in the United States. Big Data & Society, 6(1), pp. 1-5.

^[13] Prof. Banafa, A. (2023). The use of AI in the criminal justice field <https://www.linkedin.com/pulse/use-ai-criminal-justice-field-prof-ahmed-banafa-mrtnf> (Last access: 20.07.2024).

^[14] Kelly, W. R. (2016). The future of crime and punishment: Smart policies for reducing Crime and saving money. Rowman & Littlefield Publishers, USA  pp. 72-74.

^[15] Rigano, G. (2018) Using artificial intelligence to address criminal justice needs. National institue of justice (NIJ) https://nij.ojp.gov/topics/articles/using-artificial-intelligence-address-criminal-justice-needs (Last access: 20.07.2024).

^[16] Carpertner, W. (2023). Presidents’ massage: AI’s transformative impact. <https://www.policechiefmagazine.org/presidents-message-ais-transformative-impact/> (Last access: 22.07.2024).

^[17] Antunes, H. S., Freitas, P. M., Oliviera, A. L., Pereira, C. M., Sequiera, E. V., Xavier, L. B. (2023). Multidisciplinary Perspectives on Artificial intelligence and the Law. Springer, Spain, pp. 15-20.

^[18] Vasdani, T. (2023). From Estonian AI Judges to robot mediators in Canada. U.K, The Lawyer’s Daily, Vol. III. pp. 5-10.

^[19] Wischmeyer, T. (2020). Artificial intelligence and transparency: Opening the black box. Regulating Artificial Intelligence, pp. 75-101.

^[20] Mugari, I., Obioga, E. (2021). Predictive Policing and Crime Control in The United States of America and Europe: Trends in a Decade of Research and the Future of Predictive Policing. Social Sciences, 10(6), p. 234.

^[21] Yucel, G. (2024). Crime prevention effort with artificial intelligence in China: analysis <https://www.linkedin.com/pulse/crime-prevention-efforts-artificial-intelligence-china-gokcen-yucel-aascf> (Last access: 25.06.2024).

^[22] Industry views. (2024). Mahalias, I. AI adoption in criminal justice – How can industry support the justice system in implementing Artificial intelligence <https://www.techuk.org/resource/ai-adoption-in-criminal-justice-how-can-industry-support-the-justice-system-in-implementing-artificial-intelligence.html> ( Last access: 25.07.2024).

^[23] State v. Loomis. (2016). Harward law review, Vol. 130. Issue 5, Harward University, pp. 1530-1537.

^[24] Hunton, A. K. (2020). UK court of appeal finds automated facial recognition technology unlawful in Bridges v South Wales police <https://www.huntonak.com/privacy-and-information-security-law/uk-court-of-appeal-finds-automated-facial-recognition-technology-unlawful-in-bridges-v-south-wales-police> (Last access: 20.07.2024).

^[25] Szappanyos, M. (2023). Artificial intelligence: Is the European court of Human rights prepared? Acta Humana, Vol. 11(1), pp. 93-110.