When Is Using Artificial Intelligence a Crime?

A judge explaining the legal limits on artificial intelligence use stressed that merely using the technology is not a crime, but if its use conflicts with the law, legal liability will be determined based on the conduct carried out.

Ambassador of Arjan-Social Group: Shamsuddin Jalilipiran, Judge of Justice and PhD in International Law in a conversation with Mizan, about whether the use of artificial intelligence is legally prohibited? He answered in detail.

In this regard, Jalili Piran said: for an accurate answer, one must distinguish between the use of artificial intelligence and the type or method of using artificial intelligence. The mere use of a technology, regardless of the application and without the existence of a legal basis, cannot be considered a crime or prohibited. On the other hand, this does not mean that every use of artificial intelligence is legal in every situation and for every purpose.

An important principle in criminal law is the legality of crime and punishment. Article 36 of the Constitution stipulates that punishment and its execution must be done only by a competent court and according to the law, and Article 37 emphasizes the principle of acquittal. Therefore, the newness of a technology or concern about its effects alone cannot be the basis for creating a criminal responsibility.

As a result, when examining the legality of an artificial intelligence application, one must ask what this technology was used for, what data was provided to it, what output or effect it created, and whether the behavior performed was in conflict with the law or not.

The legality of the principle of using technology does not mean that every way of using it is legal

Artificial intelligence may be used in areas such as education, research, translation, data analysis, programming, content production or editing, and professional activities. But the legality of the principle of using technology does not mean that every way of using it is legal. If during the use of artificial intelligence, the rights of individuals, confidential information, protected data, intellectual property rights or legal and contractual obligations are violated, the same specific behavior should be investigated according to the relevant regulations.

Simply using artificial intelligence is not a crime

This distinction is more important in the criminal field. If someone uses artificial intelligence to commit a behavior that is criminalized by law, the use of new technology does not decriminalize the behavior. For example, if an artificial intelligence tool is used in a process to facilitate unauthorized access to data or computer systems, the matter should be investigated in light of the conduct and the provisions relating to computer crimes; Not that the mere use of artificial intelligence is considered a crime.

What does Article 1 of the Computer Crimes Law say?

For example, Article 1 of the Computer Crimes Law criminalizes unauthorized access to protected computer or telecommunications data or systems; Therefore, if artificial intelligence technology in a process helps to commit such behavior, the criminal description comes from the criminalized legal behavior, not from the title of using artificial intelligence.

A distinction should be made between crime and other forms of violation of the law

On the other hand, we must differentiate between crime and other forms of law violations. Any breach of privacy, any contractual dispute, any intellectual property issue, or any misuse of data is not necessarily a criminal offense. A conduct may give rise to civil, contractual, administrative or professional liability as the case may be, without having the elements of a criminal offence.

For this reason, in the face of artificial intelligence, it is better to talk about the application, the context of use, the data used, the effects of behavior and the regulations governing it, instead of proposing a general ban. A use may be permitted in one area, but the same technology may face restrictions or legal liability in another area by changing the purpose or manner of use.

EU AI regulation for high-risk AI systems

In sensitive areas, the issue is not only whether the use of technology is allowed or prohibited; Moreover, the limits of authority, the way of monitoring, the ability to check the output, the possibility of modifying it and appointing the responsible person also become important. In this context, for example, the European Union's Artificial Intelligence Regulation for high-risk artificial intelligence systems foresees specific requirements about human supervision.

Article 14 of this regulation emphasizes the possibility of effective monitoring of natural persons, recognition of the capabilities and limitations of the system, attention to the risk of excessive reliance on the output of the system and the possibility of ignoring, modifying or stopping it. This regulation is not the law of Iran and is merely a comparative example of the legal regulation of artificial intelligence in the European Union.

No technology is outside the realm of the law

Therefore, in Iranian law, the mere use of artificial intelligence, regardless of the type of use and without the existence of a legal basis, cannot be considered a crime or prohibited; But no technology is outside the realm of the law. If the way of using artificial intelligence is in conflict with the law, the responsibility should be determined based on the same behavior and related regulations, by meeting the legal requirements.