ARTIFICIAL INTELLIGENCE

AS A HUMAN SERVANT OR VICE VERSA:

LEGAL ASPECTS

УДК: 347.73

ARTIFICIAL INTELLIGENCE

AS A HUMAN SERVANT OR VICE VERSA:

LEGAL ASPECTS

Kildeev Adel

Zhdanov A. Leningrad State University (1990)

Philosophiæ Doctor in Law

ORCID: 009-0001-2211-4944

Abstract: This article examines the problems of artificial intelligence and its legal status, analyzes various positions of legal scholars on this issue and legislative regulation. Objective: to consider in a critical aspect scientific approaches to determining the legal status of artificial intelligence (AI) both in our country and abroad. Methodological basis: the method of comparative analysis. Results: Russian and foreign concepts of the legal status of artificial intelligence are considered. Conclusions: AI and robots are objects of law, but not subjects, attempts to give them the status of an individual, animal or legal entity are unreasonable and ridiculous. To determine the status of AI and robots, the status of a slave, enshrined in the basis of civil law – Roman law, is best suited, of course, subject to a certain modification of it. 

Keywords: Robots, robotics, artificial intelligence, intellectual property, copyright, legal personality, legal status, Roman law, slave, court.

The topic of artificial intelligence and its application in all spheres of life is one of the most relevant now. The Internet is filled with photos and videos generated by artificial intelligence (AI or AI in English transcription), which are no longer inferior in quality to professional photos and clips. Artificial photos, videos, songs, lyrics, translations of texts generated by AI – this flurry has hit our consciousness quite recently. Are we ready for AI and will we be able to control artificial intelligence? Or will the gloomy predictions of science fiction writers come true and we will face the dark times of slavery under the heel of machines?

Stargate will be created in the United States, whose task will be to develop artificial intelligence infrastructure, US President Donald Trump announced along with the heads of OpenAI Sam Altman, SoftBank Masayoshi Son and Oracle Larry Ellison. According to Trump, Stargate will be “the largest AI infrastructure project in history”; will build “physical and virtual infrastructure to support the next generation of artificial intelligence,” including data centers across the United States. The first data center is already under construction in Texas. The three companies will invest $100 billion in the project, investments will increase to $500 billion in the coming years, and 100,000 new jobs will be created in the United States. Artificial intelligence requires a huge amount of infrastructure, power, computer chips, data centers, Sam Altman said in a recent interview, so it is necessary to create all this in the United States in order to maintain leadership in technology.^[1]

On January 20, 2025, the Chinese company DeepSeek released the DeepSeek R1-Zero and DeepSeek R1 models. The DeepSeek neural network is based on a large open-source language model and is available in Russian. “In January 2025, DeepSeek unveiled its flagship model, the DeepSeek-R1, which is comparable in capabilities to cutting-edge developments such as OpenAI’s GPT-4o and o1. However, its training was much cheaper – about $6 million, which is several times lower than the cost of GPT-4, the development of which in 2023 required $100 million,” comments Kirill Pshinnik, co-founder and CEO of the Zerocoder online university.^[2]

On January 30, 2025, Alibaba Group Holding Ltd. unveiled an updated Qwen 2.5 Max AI model, stating that it outperforms Meta* (Llama) and DeepSeek (V3) solutions in key performance tests. According to the official statement, the company’s new development has shown a world-class level in a number of independent tests.^[3]

There is a struggle of global corporations for the consumer, more and more new AI models will be released on the market, more and more advanced and cheaper. Shares of AI chip manufacturers are collapsing, investors are at a loss: which technologies are more profitable to invest in?

Governments around the world understand the need to regulate AI. At the AI Action Summit in Paris, which ended on February 11, 2025, 58 countries signed a declaration on the “open”, “inclusive” and “ethical” development of AI (Russia did not participate in the Summit for obvious political reasons). The Summit was a significant event, bringing together representatives of states, international organisations and leading companies to discuss the future of artificial intelligence. The event was organised by France and India, which called for global cooperation in the field of AI.

One of the important points of the declaration was the commitment to “make artificial intelligence sustainable for people and the planet.” In this regard, the summit announced the creation of the Observatory of the Energy Impact of Artificial Intelligence, which will operate under the auspices of the International Energy Agency (IEA). The Coalition for Sustainable AI was also established, which will include leading companies in the sector.

However, the UK and the United States did not sign the declaration. On the UK side, Prime Minister Keir Starmer’s spokesman explained the refusal to sign, citing national interests. “We only join initiatives that are in the interests of the UK,” he said. U.S. officials did not provide official reasons for not signing the declaration. But from the rostrum of the Paris summit, Donald Trump’s administration’s special envoy, Vice President J.D. Vance, criticized excessive regulation of AI, saying that “over-intervention could kill an emerging industry.” The refusal of the United States and the United Kingdom to sign the declaration illustrates the existing differences in approaches to regulating AI.

Let’s go back to the roots. Robert Heinlein, Isaac Asimov, Philip K. Dick, Stanislaw Lem, Arthur C. Clarke, the Strugatsky brothers – this is an incomplete list of science fiction writers who considered the problems of coexistence between humans and artificial intelligence in their works.  It should be noted that the writers have thoroughly studied almost all aspects of such interaction from the point of view of psychology, sociology, and cybernetics. Hollywood, European cinema, and our own have repeatedly exaggerated this topic, giving either an idealistic and tearful approach (the films “Artificial Intelligence” or “Bicentennial Man”), or a harshly post-apocalyptic one (the Terminator franchise film), or a comedy one (the series “Love, Death + Robots”, “The Anna Nikolaevna Project”, “Better Than People”).

And what do lawyers think about AI? Someone considers the applied aspects of AI borrowing other people’s works as copyright infringement, quite reasonably believing that “since artificial intelligence is not a source of new content, but only serves as a tool in the hands of the user, then in fact it is the user who infringes these copyrights.”^[4]

Someone is already actively using AI in jurisprudence and promoting its further application^[5].  Canada and Quebec have gone further than many countries in regulating AI and defining ethical and legal norms for its use^[6], limiting the use of AI in the judicial system in some way^[7]. Moreover, the Federal Court of Canada has banned its judges from using AI in decision-making^[8].  

And how do Russian lawyers see the problem of the legal status and regulation of AI? A detailed analysis of this issue is given in the article “Legal Status of Artificial Intelligence” by A.V. Afanasyevskaya, Ph.D.: “Several concepts on this problem have been formed in domestic legislation. The first of them proposed to equate the status of AI with the status of an individual, thereby extending to him the entire range of rights and obligations regulated by the current legislation. Supporters of the second concept believe that it is necessary to equate the status of robotics with the civil status of animals. The third concept is based on the idea that proposes to use the analogy of legal regulation of the status of a legal entity in civil law relations to solve this problem. Today, the most acceptable in terms of features and specifics, according to the domestic legislator, is to equate the status of AI with the status of a legal entity. For these purposes, the legislator has developed a package of amendments, it is proposed to include in the Civil Code of the Russian Federation a special chapter devoted to these issues, with the following wording: “A robotic agent (robot agent) is a robot that, by the will of the owner and by virtue of design features, is intended to participate in civil turnover. This agent has separate property and is responsible for all his obligations, can acquire and exercise civil rights and bear civil obligations on his own behalf, be a plaintiff and a defendant in court.”^[9]

Similar positions are set forth in a number of works by other lawyers-scientists, for example, in the article by I.R. Begichev and Z.I. Khisamova “Artificial Intelligence and Robotics: Theoretical and Legal Problems of Differentiation of the Conceptual Apparatus”. The authors, highlighting in great detail the features of a robot and AI, rightly note that “other concepts (“robot as an animal”, “robot as a person”, “robot as a legal entity”) also have a number of shortcomings, including, in particular, complication of legal regulation, problems with rethinking the basic values of society, an attempt to evade responsibility by “replacing the object”, etc. ^[10]Similar points of view are contained in the article by Doctor of Law A.V. Popova “New Subjects of the Information Society and the Society of Knowledge: On the Issue of Legal Regulation”, which believes that “a robotic device or an industrial robotic line is an unconditional object of civil law relations, including intellectual property rights, but an industrial robot, a robot agent, a robot serving a robot, depending on the use of AI in it, can also be a special subject of law with limited legal capacity”^[11]. Interesting aspects of this complex problem are raised in the article by D.Y. Mikhailov and Doctor of Law A.V. Popova “Legal Regulation of Artificial Intelligence – Dehumanization of Law?”: “If some people still do not have full legal capacity due to sexual, racial or other discrimination, then the question of whether an electronic person should have legal capacity, independently make decisions and be responsible for them is relevant.”^[12]

In my opinion, when determining the legal status of AI, it is time for us to return to the basics of our civil law – Roman law. 

A slave (servus) was outside of political society and was not a subject of law, according to Roman law he was considered a thing. In the sphere of property, the slave acted as an instrumentum vocale (“speaking instrument”). The ability of a slave to perform legal actions (but only in favor of the master) was derived from the legal capacity of the latter, that is, as if such actions were performed by a slave-owner, while the master did not bear any responsibility for the obligations of slaves. was modified, and slaveholders became liable for the obligations of slaves arising in the process of carrying out their economic activities on behalf of their masters. Peculium is a separate property allocated to a slave by the master for independent management, part of the income from which the slave had to contribute to the master. Since the economic activity was carried out by the slave on his own initiative and in his own interests, and the peculium legally belonged to the master, he was responsible for the obligations of the slave within the boundaries of the peculium^[13].

Of course, the full application of the doctrines of Roman law to the status of AI is impossible without their modernization, but the main thing in the doctrine is that a slave is considered a thing, just as AI is a thing, and not a person or an animal, which some researchers write about completely thoughtlessly. I am afraid that they would never have been able to pass the Roman law exam to my favorite professor V.A. Musin^[14] with such doubtful knowledge. Well, what legal capacity can a “smart” speaker or iron have? What personal rights and obligations can a toaster or refrigerator have? What is the legal responsibility of a copier? Why are a donkey and a smart electric car animals? How to impute rape to a vibrator? What are the alimony obligations of the meat grinder?

We live in a changing and crazy world where dehumanization is in full swing. The largest transnational corporations promote their latest technologies that displace people from production and from life in general. It is more profitable to pay a person unemployment benefits than to provide him with a skilled job. Because his work is already occupied by a robot AI. According to analysts’ forecasts, in the next 10-15 years, about half, if not more, of jobs will be lost to people, the work will be done by robots, the place of writers, translators, journalists, teachers, actors will be taken by AI, robotic doctors will treat people, and robotic lawyers will sue among themselves. I am exaggerating a little, of course, but the general picture of the future world, outlined by the world backstage (oligarchs), is already quite visible. We must ask ourselves the question: why do we live in this world and what do we strive for? What is the main value? A person and his happiness or making a profit by any means with the accumulation of these capitals in the hands of a handful of unscrupulous oligarchs? The illusion of AI controllability has already begun to fade, and numerous cases of machine disobedience are no longer a secret. The same applies to the dehumanization of law, the admission of AI into the process of lawmaking, judicial activities and the execution of court decisions is completely unacceptable, in my opinion, at least for the time being.

In conclusion, I would like to cite an excerpt from my favorite fantasy story by Gordon Dixon, “Punch Cards Don’t Discuss”, in which AI, by its mistake and organic idiocy, sentences to death a reader who did not turn in Robert Louis Stevenson’s book “Kidnapped” to the library on time:

“The State of Illinois.

Extremely urgent!

I, Hubert Daniel Willikens, Governor of the State of Illinois, who by virtue of the power given to me to pardon persons who, in my deep conviction, have been convicted too cruelly and deserve pardon, hereby pronounce the pardon of Walter A. Child (A. Walter), who is now in prison as a result of a wrongful conviction for a crime of which he is absolutely innocent, absolving him of all such guilt. According to this pardon, Walter A. Child (A. Walter) is to be released from custody…

Interdepartmental Relations Service.

Please do not bend, crumple or fold the punched card.

Content: Failure to comply with the rules for sending documents.

Where: In the office of Governor Hubert Daniel Willikens.

Regard: “Pardon of Walter A. Child on July 1, 1987.”

Dear Mr. Chief of Staff,

You forgot to put the outgoing number on the said document. We ask you to re-transfer this document to the punched card on Form 876, the only one that allows you to put the designation “extremely urgent” on the document. Form 876 must be signed by your superior.

The Bureau of the Interdepartmental Liaison Service was opened on Wednesday, 5 July 1987.

Warning: If you fail to submit your Form 876 document, you may be prosecuted for non-compliance.”^[15]

Do you want to live in such a world? I don’t! And think about why on all submarines, airplanes and spacecraft, command and control systems are repeatedly duplicated, and if possible mechanically.

References

1. Afanasyevskaya A.V. Legal status of artificial intelligence. / Bulletin of the Saratov State Law Academy, No 4 (141) 2021, p.90.

2. Begishev I.R., Khisamova Z.I. Artificial intelligence and robotics: theoretical and legal problems of differentiation of the conceptual apparatus. Bulletin of the Udmurt University, vol.30, issue 5, 2020, p.711.

3. Grimm, D.D. Lectures on the Dogme of Roman Law (Fifth Edition, St. Petersburg, 1916). Reprint. Moscow, Zertsalo Publishing House, 2003.

4. Dixon, Gordon. Punch Cards Do Not Discuss / Electronic Library LitMir https://litmir.club/br/?b=58580

5. Artificial Intelligence and Copyright – Who Will Win?/ An article without the author’s name on the website “IT Lawyers | SHEVTSOV.RU

https://shewzov.ru/articles/artificial-intelligence-and-copyright-who-will-win/

6. Mikhailov D.Yu., Popova A.V. Legal Regulation of Artificial Intelligence – Dehumanization of Law? Education and Law No 8, 2020, p.269.

7. DeepSeek neural network: what it can do, how to use, a description of the artificial intelligence model / RIA Novosti. 20:24 31.01.2025.

Electronic resource: https://ria.ru/20250131/deepseek-1996714057.html

8. Popova A.V. New Subjects of the Information Society and the Society of Knowledge: On the Issue of Normative Legal Regulation. Zhurnal Rossiyskogo Prava, No 11, 2018.

9. Trump announced the largest AI project in history | Open Systems Publishing House.

Electronic resource: https://www.osp.ru/articles/2024/0127/13059137

10. Alibaba claims superiority of its new AI model over DeepSeek and Meta* / AI, services and apps / iXBT Live. Electronic resource: https://www.ixbt.com/live/sw/alibaba-zayavila-o-prevoshodstve-svoey-novoy-ai-modeli-nad-razrabotkami-deepseek-i-meta.html

11. Federal Court bans its judges from using AI in decisions in wake of U.S. controversy / The Globe and Mail, January 9th 2024.

https://www.theglobeandmail.com/canada/article-ai-canada-federal-court-judges/

12. Generative AI in the Legal Profession / Harvard Law School Center on the Legal Profession, 2022

https://clp.law.harvard.edu/knowledge-hub/magazine/issues/generative-ai-in-the-legal-profession/

13. Guidelines for the Use of Artificial Intelligence in Canadian Courts. Guidelines for the Use of Artificial Intelligence in Canadian Courts. Prepared by Martin Felsky, Ph.D., J.D., and Professor Karen Eltis for the Canadian Judicial Council First Edition, September 2024.

https://cjc-ccm.ca/sites/default/files/documents/2024/AI%20Guidelines%20-%20FINAL%20-%202024-09%20-%20EN.pdf

14. Legal framework for artificial intelligence: Where do we stand in Canada and Quebec? November 10th, 2023| Langlois Lawyers

https://langlois.ca/en/insights/legal-framework-for-artificial-intelligence-where-do-we-stand-in-canada-and-quebec

1. Trump Announces Largest AI Project in History | Open Systems Publishing. Electronic resource: https://www.osp.ru/articles/2024/0127/13059137 ↑

2. DeepSeek neural network: what it can do, how to use it, a description of the artificial intelligence model / RIA Novosti. 20:24 31.01.2025.

   Electronic resource: https://ria.ru/20250131/deepseek-1996714057.html ↑

3. Alibaba claimed the superiority of its new AI model over the developments of DeepSeek and Meta* / AI, services and applications / iXBT Live. Electronic resource: https://www.ixbt.com/live/sw/alibaba-zayavila-o-prevoshodstve-svoey-novoy-ai-modeli-nad-razrabotkami-deepseek-i-meta.html ↑

4. Artificial Intelligence and Copyright – Who Wins?/ An article without the author’s name on the website “IT Lawyers | SHEVTSOV.RU https://shewzov.ru/articles/artificial-intelligence-and-copyright-who-will-win/ ↑

5. A Guide to the Use of Artificial Intelligence in Legal Practice: Generative AI in the Legal Profession / Harvard Law School Center on the Legal Profession, 2022 https://clp.law.harvard.edu/knowledge-hub/magazine/issues/generative-ai-in-the-legal-profession/ ↑

6. Legal framework for artificial intelligence: Where do we stand in Canada and Quebec? November 10th, 2023| Langlois Lawyers https://langlois.ca/en/insights/legal-framework-for-artificial-intelligence-where-do-we-stand-in-canada-and-quebec/ ↑

7. Руководство по использованию искусственного интеллекта в канадских судах (англ.): Guidelines for the Use of Artificial Intelligence in Canadian Courts. Guidelines for the Use of Artificial Intelligence in Canadian Courts. Prepared by Martin Felsky, Ph.D., J.D., and Professor Karen Eltis for the Canadian Judicial Council First Edition, September 2024. https://cjc-ccm.ca/sites/default/files/documents/2024/AI%20Guidelines%20-%20FINAL%20-%202024-09%20-%20EN.pdf ↑

8. Federal Court bans its judges from using AI in decisions in wake of U.S. controversy / The Globe and Mail, January 9th 2024. https://www.theglobeandmail.com/canada/article-ai-canada-federal-court-judges/ ↑

9. Afanasyevskaya A.V. Legal Status of Artificial Intelligence. / Bulletin of the Saratov State Law Academy, No 4 (141) 2021, p.90. ↑

10. Begishev I.R., Khisamova Z.I. Artificial Intelligence and Robotics: Theoretical and Legal Problems of Differentiation of the Conceptual Apparatus. Bulletin of the Udmurt University, vol.30, issue 5, 2020, p.711. ↑

11. Popova A.V. New Subjects of the Information Society and the Society of Knowledge: On the Issue of Normative Legal Regulation. The Journal of Russian Law, No 11, 2018, p.23. ↑

12. Mikhailov D.Yu., Popova A.V. Legal Regulation of Artificial Intelligence – Dehumanization of Law? Education and Law No 8, 2020, p.269. ↑

13. Grimm, D.D. Lectures on the Dogma of Roman Law (fifth edition, St. Petersburg, 1916). Reprint. Moscow, Zertsalo Publishing House, 2003. ↑

14. Musin, Valery Abramovich (1939 – 2015) – lecturer at the Faculty of Law of Leningrad State University, Doctor of Law (1979), Professor (1982). ↑

15. Dixon, Gordon. Punch Cards Do Not Discuss / Electronic Library LitMir https://litmir.club/br/?b=58580 ↑