Criminal Liability Of Artificial Intelligence Entities
Author: LawSphereX
Category: Criminal Law
Published:
Updated:
CRIMINAL LIABILITY OF ARTIFICIAL INTELLIGENCE
ENTITIES
Jayansh Sharma, Manipal University Jaipur
ABSTRACT
The evolution of information technologies has led to the development of artificial intelligence and it has led us to a point where we are face-to-face or confronted with another entity that is fully capable of functioning on its own without any human intervention, means it can be said that they have their mind that makes them capable of taking decision and to choose between what is right and what is wrong, and in the world of this modern technologies the rise of artificial intelligence has changed various sectors and as a result of this there are various complex legal issues and the main question of the accountability of the criminal liability.
The invention of artificial intelligence can be seen as a venture to look into the human minds as to interpretation and decision-making power of human beings, however, the capability of acting without any human intervention can harm the interest of any individual or collective group of people. This paper aims to emphasise the points of applicability of the concept of criminal liabilities to artificial intelligence entities, taking account of the capabilities of the artificial intelligence of autonomous decision-making powers and what are the possible punishment that can be awarded or carried out as to put a deterrence effect on the crimes that are carried out by the artificial intelligence and what are the possible harm that can be caused by such entities.
The main purpose of the study is to determine the legal obligation that is binding upon artificial intelligence and, if there is any crime committed, whom to hold liable for such act as to define exactly the point where the inventors are exempted from the act of artificial intelligence.
Keywords: Criminal liability, artificial intelligence, accountability, punishment DEFINITION There is no particular definition that can limit the scope of the artificial intelligence but it can be said that the artificial intelligence is the study of the human nature and the attempt to understand the limits of the human intelligence and the functions of the brain, means where the same work that has been traditionally done by human beings when the same work is done by using a software through the artificial intelligence on the basis of the input that is provided by the user of such artificial intelligence1.
INTRODUCTION
The concept of artificial intelligence is the product of the human efforts, means the creator of the artificial intelligence is a human being, and it has been one of the biggest achievement of the 21st century and it has changed more of brought a revolution in sectors like finance, transportation and health care and enforcement of law, as the time goes by the percentage of human intervention in the functions of the artificial intelligence is reduced and it has led to the complex legal question of the accountability of criminal liability, and on the contrary of the point the human is more dependent on the artificial intelligence rather than the artificial intelligence used to be on the human beings.
There is hardly any domain that is left out of touch from the branches of the artificial intelligence; the artificial intelligence has made the life of human beings very easy and convenient as it aims to save time and energy2. There is no particular definition for the artificial intelligence but it can be divided in two types:- The first one is the narrow artificial intelligence that is formulated to do a particular or specific function or with any particular object, it is designed to perform specific function , it is the most common type of the artificial intelligence, and it is for the benefit of the public domain and on the other hand the second type of artificial intelligence is known as artificial general intelligence that possess the capability of understanding things and perceiving knowledge and application of such knowledge in various public domain.
The current legal system that is followed in India has minimal laws that are applicable on this particular situation and there are hardly any cases on the artificial intelligence that determines 1 Abanti Bose, Artificial Intelligence and the shift in liability iPleaders (2022), https://blog. ipleaders. in/artificialintelligenceshiftliability/#:~:text=The%20two%20possible%20results%20are,it%20shall%20be%20held%20liable. (last visited Nov 18, 2024). 2 'O. E. R. adutniy, Criminal Liability of the Artificial Intelligence, 138 Probs. Legality 132 (2017). the accountability of criminal acts3.
METHODOLOGY
The current paper uses a method in legal research that is known as doctrinal methodology that aims to investigate and provide support to above mentioned contention, the following method provided support to above mentioned contention by using data sources to fulfil the object of the study, along with the credit has been given where any information or any static has been used from any credible source to maintain the integrity of the scholar and the original work. The methods aims to evaluate the existing legal framework and principles and if there is any case laws to support the contention, it also aims to take note of legal texts, opinion of judicial mind and the writing of the scholars on the above particular domain as it is one of core elements to draw any conclusion and any legal arguments and forming of the legal policies on the artificial intelligence.
GENERAL ELEMENTS OF CRIMINAL LIABILITY There are certain elements that needs to be fulfilled in order to establish criminal liability of any individual, first the act needs to be prohibited by the law, in order to establish the criminal liability for a act, the act needs to be forbidden by any statue for the time being in force and provided that the same act should also be punishable by the law, means that such act should be backed by sanctions provided by the state. The main factors that need to be fulfilled in order to establish criminal liability is the physical element(actus reus) and the guilty mind (mens- rea), the physical element “Actus reus” reflects the wrongful act or omission and the other denotes the guilty mind.
The guilty mind reflects the motive, intention and Knowledge, provided that the certain act that falls under negligence and strict liability are the exception of the following rules, means in order to held a person accountable for a criminal act both the physical and mental element needs to be present at the same time, not having one element and having the other element does not make person accountable for the liability of the criminal act, provided any entity whether existing in physical form and artificial form satisfies the above mentioned element in that case it will attract the criminal liability, but the catch is that the both the elements needs to be present there at the same time in order to establish the liability.
3 Cogent Social Sciences 2024, Vol. 10, no. 1, 2343195
The following legal principle of “Actus Non Facit Reum Nisi Mens Sit Rea” underlines or established the same contention that mentioned above, means an act does not make a person guilty unless there is a guilty mind4. ARTIFICIAL INTELLIGENCE INVOLVEMENT IN CRIMES As we are going towards the modern technology and industrialization the needs of human beings is not limited to necessities only in the greed of getting more and more has lead to the development of cyber crimes that are committed by using artificial intelligence and such problem has become one of the major concern in the legal landscape, means the criminal mind uses artificial intelligence algorithm to further their cause for example hacking into the data base or having the personal information of any person in order to cause loss to any interest of such person.
One of the most challenging situation for the law enforcement agencies is to trace the criminals, the criminals are using advanced algorithm to mitigate their chances of being traced because there is no need of physical presence at the time of committing the crime because of such cause it has become impossible for the law enforcement agencies to trace the origin or identify the actual culprit behind such organized crime, provided there is also ambiguity in the area as to accountability of such act as it is complex to decide whether the act committed by the artificial intelligence was self generated or it was instructed to do so. The evolving use of the artificial intelligence can be used for breach in the private data and security of a person because by having access to such private data by using artificial intelligence can lead to misuse and compromising the safety of the individual.
The current legal framework should adopt the measures to prevent the breach of any personal data and information in order to secure the integrity and safety of the individual. The contention of autonomous weapons runs parallel to the development of artificial intelligence, the idea of autonomous weapons and use of such weapon without any human intervention is concerning as it can lead to loss of human life and resources. The legal framework needs to foster the innovative ideas of artificial intelligence along with measurement to safeguard the citizens along with data protection and safety5. 4 Baron, m. (2020). negligence, men’s rea, and what we want the element of mens rea to provide. Criminal Law and Philosophy, 14(1), 69–13 5 gunawan, y. , aulawi, m. h. , anggriawan, r. , & Putro, t. a. (2022). Command responsibility of autonomous weapons under international humanitarian law.
Cogent Social Sciences, 8(1), 2139906 POSSIBLE SITUATIONS IN ROLE OF ARTIFICIAL INTELLIGENCE 1. Innocent Agent - The first possible outcome in this situation is where the artificial intelligence is working on the instruction that is provided by the user, in that case the artificial intelligence works upon program or instructions provided by the user and the user intentionally provides instructions that leads to rise of criminal liability. For example, “a” programs his robot to torch the house of his enemy; the input given by the user gives rise to criminal liability. 2. Semi-Innocent Agent - The second outcome of this situation is based upon the foreseeability of the user to a potentially commission of the offence; in this particular situation, the user of the artificial intelligence works very closely; in this situation, the crime is not intended. 3.
Artificial intelligence working as individual - Where the artificial intelligence works on its own, the functions of the artificial intelligence are working without any human intervention, in this kind of situation the liability of such criminal act cannot be attributed towards the user of the interface, the artificial intelligence works independently without any input given by the programmer.
CHALLENGES
IN DETERMINING CRIMINAL LIABILITY In order to make a person of guilty of criminal liability the element of actus reus and mens rea, means the guilty mind and the act or omission along with such state of mind needs to be present there, there are various methods that is used upon the criminals to determine their mental state at the time of commission of the offence, but the same cannot be applied when it comes to the artificial intelligence because it is a software program it does not have any physical attributes or structure, it works mainly upon the instructions that are provided by the user of such program and as the artificial intelligence becomes more sophisticated and advanced the main question that who should be held accountable for the act becomes more relevant.
The legal framework in many of the countries applies upon the legal personhood means, and it is referred to individual, corporation business entities where the person managing the affairs of the business can be held liable for any criminal offence but provided that person should have a guilty mind and has done an act or omission. The provided principle of legal personhood does not apply on the artificial intelligence; also the artificial intelligence can work on its own without any human intervention.
The rapid advancement and development in the field of artificial intelligence further complicates the situation where it is not defined till which extent the human being or the user or the programmer of the artificial intelligence is liable for the actions of the artificial intelligence because there is very thin line in this complex issue that can attract the liability of any criminal offence and an individual can be held accountable for criminal liability arising out of the actions of the artificial intelligence. The lack of transparency affects the capability of the establishing the criminal liability.
Moreover the current legal framework the Indian penal code or otherwise known as the Bhartiya nyaya sanhita does not take the possibility or the scope of the current legal framework is limited to the possibility of a human being committing an offence, as a result of this the provision of the current legal framework does not apply to the artificial intelligence. In order to determine the liability of the artificial intelligence for an criminal act, first the scope that the artificial intelligence falls under the ambit of the legal entity is to be understood, as the scope of legal entity is ambiguous it becomes next to impossible determining the liability of artificial intelligence in order to held accountable.
PRINCIPLE OF FAIRNESS AND ETHICS Apart from the legal framework that is established for determining the criminal liability, there are certain principles that are embodied in the law making policy and one of the such principles is the principle of fairness and ethics along with the principle of equity, these principles are of utmost importance in determining the criminal liability, as the scope of the artificial intelligence also extends to the social functions and legal framework, means the process where the liability for certain act is determined or attributed it should take notice of the principle of fairness and ethics. The current situation needs critical examination to uphold the legal principles while determining the liability.
PROPOSED RESTRICTION ON ARTIFICIAL INTELLIGENCE The proposed measures for putting restriction on the unlimited power that has been bestowed with the artificial intelligence seek to address the complex issues that have been brought up by the rapid growth of the artificial intelligence; there is clear need of establishing the required legal framework. One of the principles that has been mentioned to propose in the strict liability, the underlined principle proposes that the person who has created or programmed such software and where such program uses artificial intelligence for its functions in that case the programmer can be held liable for the actions of the artificial intelligence, because such person can be held as a legal entity and criminal liability can be associated with such person for programming such function.
This approach explains or suggests that it is the utmost duty of the creator or the programmer for the artificial intelligence that such artificial intelligence should be programmed in that manner where any misuse of powers that have been bestowed upon the artificial intelligence is not possible; to support such cause all the detailed documentation and regular audit and inspection of such program is required, there should be transparency measures in building of such program, so the public can have access to required information and trust upon such software.
Furthermore, there is dire need of a comprehensive legal framework and further measures to ensure effective implementation of such legal framework, one of the main issues that has to be addressed in these provisions is the safety and protection of the private and personal data of the end consumer of such software, where the application or the software requires personal information in order to effective use of such program it is the utmost duty of the developer of the program to keep the provided data safe and maintain the integrity of the data, the legal framework can ensure that such personal data of the end consumer is not used or harvest for any criminal activity, where it can harm the interest of the end consumer and such data can be accessed by the other person without the prior consent of the person whom with such data is associated.
One of the proposed measurements for such cause is the recognition of crimes that is related to artificial intelligence, taking note of the rapid development of the artificial intelligence it is in better interest that emerging crimes should be recognized, there is need of transparency in the process of the creation of the artificial intelligence, the required information should be easy to access and should be available in public domain. INTERNATIONAL ASPECT OF ARTIFICIAL INTELLIGENCE Our day to day aspects have been changed by Artificial Intelligence (AI) over the period of time, from enhancing healthcare to jobs being automated. As there is more degree of involvement of AI in the technologies we use everyday, there arises important question as who should be responsible when AI malfunctions or causes harm. Unlike us i.
e, humans the AI cannot think and feel instead the decision made by them are based on data and programming.
Since there is lack of emotions and intentions in AI, it is very difficult to certain that when something wrong occurs whom should be held liable. That’s why there is serious need for legal frameworks which can explain who shall be held liable when AI causes any harm. The issue of AI liability is approached in a different way in different countries. The European Union, United States and India do handle the query of AI responsibility in their respective manner. By learning these differences it would be beneficial for us to see how the rules are being created by countries to cope up with the new challenges which this AI brings. Approach By United States’: Using existing laws AI in the United States of America being treated just as like a tool or a product, not as an living being who is independent.
So whenever any harm is caused by the AI, the person or a corporate body associated with it, are held responsible, not the AI itself. Lets take an example, if the AI used by a company for hiring decision and the end result of that decision turns out to be biased or discriminatory, such company can be blamed for the inappropriate us of AI. If a data is breached by AI system, the accountability would of the company who designed or used the AI. Due to the attributes of AI being unable to think or feel it can be difficult to figure out where the fault lies when there is some problem occurs. The main focus in United States is whether the person by whom such AI is created or used has acted responsibly. The company or developer could be made responsible if the AI was designed inappropriately or poorly.
Approach By European Union: Specific Laws for AI By creating particular legal frameworks to deal with AI and its potential risks associated, the approach by European Union (EU) can be said to be more structured. AI Act is said to be one of the notable regulations in the EU which classifies AI system by the measurement of risk they pose. For example, AI in the field of healthcare or transportation is considered to be of high risk and is subject to tight regulations. It has been assured by AI Act that AI system in these area meet high standards for reliability, safety and transparency. AI Liability Directive has been introduced by EU along with the AI Act. It has been made easier for the people to prove that the harm was caused by AI.
It mandates company to provide clear information regarding their AI system work, which assists people learn where thing went not right if there is any harm by AI. If any damage is being caused by AI both the company used and person developed such AI are held liable.
In comparison to United States the EU’s regulations are more comprehensive however, it is the complexity of AI systems which still makes it very difficult to learn how are the decisions are made by AI. Due to the complexity of the AI systems despite of these rules in existence finding out exactly where the responsibility lies can still be testing. Approach By India: Still catching with AI Laws India is still in the stage where it is still trying to find a way to regulate AI. At the moment, the country does not have any particular laws related to AI instead realies on the older technology laws, like Information Technology (IT) act of 2000. The AI is not particularly addressed by these laws hence, when there is any harm which is caused by the AI the human associated with the technology is held liable just like in the United States.
There is a law called Consumer Protection Act of 2019 in India which can be used to make companies liable if there is any fault in their AI products. However, there are no such legal frameworks which directly deals with harm or crime caused by AI itself. Rather the main focus is on whether the is any of irresponsibility by the person behind the AI. The stage of India is said to be very early in regards for the regulation of AI, and as there is evolution of the AI overtime there developed requirement for more particular laws to address AI-related risks and ensure it is used safely. The Concept of AI Legal Personhood: The idea of granting AI legal personhood is under discussion by some of the countries like Japan and Saudi Arabia. This simply means that an AI would be treated just as like people under the law.
They could own property, sign contracts or even be held accountable for their actions. However, AI in some of the countries like India, European Union, and United States is not treated like a person. In the place of this the AI is looked as a tool created by humans, which simply means humans are liable for the actions done by AI. The grant to AI of being legal personhood is a big idea which raises some questions. If AI were looked like a person, It could change how responsibility is assigned when thing don’t go right. But currently, it has been agreed upon by many countries that AI shall be treated as a tool created by human beings, and in case of any harm being caused by AI the human actors shall be responsible.
Comparing the Approaches: US, India, and EU: The clear differences can be seen while comparing the approaches to AI Liability in the United States, European Union and India: In the United States, for addressing the AI issues the law mostly uses existing rules. This approach is said to be simpler but it is less particular, so it make it harder to figure out who is responsible when AI causes harm. The European Union has a more comprehensive approach. There are laws created like AI Act and AI Liability Directive which mainly focuses on AI’s risk and require transparency. Since, AI systems can be difficult to understand the laws can be complex despite of being clear and specific. While India is still on its way to catch up. There are no specific laws for AI yet and the its reliance on older laws don’t address AI’s unique challenges.
As a result, India has to face hardship in managing the risk posed by AI and ensuring that developers and companies are held accountable. As AI continues to grow, it is essential for all nations to work together and establish legal frameworks that can resolve the risks AI poses. By framing clear and specific laws, nations can ensure that AI is used safely and with responsibilities and that those who are responsible for AI’s actions are held liable for criminal activities.
CONCLUSION
As this generation is reaching towards industrialization and modern technology, the development in technologies and the artificial intelligence cannot be avoided, as a matter of fact the invention of artificial intelligence has made life easy and convenient for the human beings, but every coin has two sides, we cannot accept the good things from the one things and avoid that harm our interest.
The concern of artificial intelligence used as tool for criminal activity is not limited to one place, artificial intelligence is used for criminal activities all over the globe, there is dire need of a structured legal framework that recognizes the offences that can be attributed to the artificial intelligence and the criminal liability should be bestowed upon the developer or the programmer who has made such software, the principle of strict liability should be recognized in these peculiar situation, the artificial intelligence should be given the legal personhood and should be recognized as entity in order to attract the criminal liability, that will maintain balance between the need of artificial intelligence and the misuse of such program