Artificial Intelligence and Law: A Legal Perspective on Digital Transformation
As digital transformation continues to facilitate our business processes day by day, it also brings along new legal risks. It can be said that artificial intelligence systems, which stand out particularly in data analysis, profiling, and optimizing production processes, are the stars of digital transformation. In addition to areas such as production and sales, the use of artificial intelligence in creative media is also increasing efficiency by enabling companies to produce advertising films or brochures, develop campaign ideas, and identify target audiences in a much shorter time. As the use of artificial intelligence in the business world grows to this extent, companies' efforts and needs to establish a legal infrastructure are also increasing in parallel.
In this article, we will first examine the structure of the system by providing a basic technical explanation of artificial intelligence, and then address the legal aspects of AI usage in companies.
A. What is Artificial Intelligence?
According to the definition provided by the Republic of Turkey Digital Transformation Office, “Artificial intelligence is the ability of a computer or a computer-controlled robot to perform tasks typically associated with intelligent beings.” In other words, artificial intelligence systems are inanimate entities produced by taking human intelligence as a model and possessing the inherently human ability to learn. To concretize this point, it should be emphasized that artificial intelligence systems, in terms of their structure and form, are essentially no different from software. However, the most significant difference between artificial intelligence and traditional software lies in its capacity for self-learning. In fact, at the current stage of technology, it has been revealed that some artificial intelligence systems learn even faster than humans.
B. Does Artificial Intelligence Learn?
Although this article will briefly touch upon how artificial intelligence systems learn (as a more accurate technical explanation should ideally be provided by computer science experts, we would like to clarify that the technical concepts covered in this article are addressed from a legal perspective, and readers should refer to technical papers for detailed explanations), a minimal level of technical knowledge will be included to facilitate an understanding of how the system works and to provide a legal interpretation.
As shown in the infographic in Figure 1, artificial intelligence is an umbrella concept encompassing machine learning and deep learning, which frequently appear in academic studies. Therefore, algorithms trained through machine learning or deep learning are collectively considered artificial intelligence in the literature. As the names of these concepts suggest, artificial intelligence systems operate based on the principle of "learning through training" and possess an autonomous decision-making mechanism rather than merely functioning as a software feature.
It is not possible to answer the question of how artificial intelligence learns from a single perspective, as there are various learning models for AI. In summary, however, artificial intelligence learns from data-driven experiences and receives feedback during the learning process to improve its models or behaviors.
Data plays the most crucial role in artificial intelligence learning. AI is trained by being fed large amounts of data and datasets to produce results requested by the user. Therefore, the more data an AI system is trained with, the more accurate its results will be. Precisely for this reason, the nature of the data inputted into AI systems, whether it is legally protected, and whether any legal violations occur during data processing are matters of utmost importance for companies utilizing AI systems.
C. Use of Artificial Intelligence in Companies and Legal Requirements
While artificial intelligence systems facilitate the operations of many companies, they also bring legal risks. Although there is no binding national or international legislation in force yet, the Artificial Intelligence Act (AI Act), which is being discussed and has a draft version approved by the European Union (EU) Commission, signals that companies will soon face certain legal obligations arising from their use of artificial intelligence.
In particular, certain obligations—such as establishing a transparent and accountable system and ensuring that the data used to train artificial intelligence systems, known as “training data,” is processed in compliance with the law—may be directly imposed on the companies that develop and even use the artificial intelligence systems.
In foreign legal systems, it has been observed that most disputes reviewed by different authorities involve data inputted into artificial intelligence systems. For example, a company that profiles its customers by feeding their personal data into an AI system must ensure that it uses personal data obtained lawfully in compliance with the relevant data protection legislation's obligations. Similarly, in generative AI systems (Generative Adversarial Networks/GAN), there is an obligation to prevent the unlawful processing of input data due to the potential copyright protection of visual or video content processed by these systems.
D. The Impact of the EU Artificial Intelligence Act on Turkish Companies
Following the official enactment of the EU Artificial Intelligence Act, not only companies originating from the EU but also Turkish companies operating within EU borders will be required to comply. Consequently, it is of utmost importance for all companies involved with artificial intelligence—especially technology firms—to expand their legal risk assessments to include AI and to incorporate this into their future strategic plans.
E. Conclusion
We would like to emphasize that regulations in the technology sector aim not to restrict companies’ activities but to regulate them in compliance with the law. Therefore, compliance with legislation does not hinder commercial activities; on the contrary, it legitimizes these activities and prevents judicial and administrative sanctions. Accordingly, every company involved with artificial intelligence systems must have the necessary legal infrastructure and ensure legal certainty in this currently somewhat gray area of regulation.
References:
Mohsen Soori , Behrooz Arezoo , Roza Dastres Artificial İntelligence, machine learning and deep learning in advanced robotics a review, Cognitive Robotics 3, 2023.
Armağan Ebru Bozkurt Yüksel, Avrupa Komisyonu’nun Yapay Zekâ Tüzük Teklifi’ne Genel Bir Bakış, Taad Dergipark, Year 13 . Issue 51 . July 2022.
As digital transformation continues to facilitate our business processes day by day, it also brings along new legal risks. It can be said that artificial intelligence systems, which stand out particularly in data analysis, profiling, and optimizing production processes, are the stars of digital transformation. In addition to areas such as production and sales, the use of artificial intelligence in creative media is also increasing efficiency by enabling companies to produce advertising films or brochures, develop campaign ideas, and identify target audiences in a much shorter time. As the use of artificial intelligence in the business world grows to this extent, companies' efforts and needs to establish a legal infrastructure are also increasing in parallel.
In this article, we will first examine the structure of the system by providing a basic technical explanation of artificial intelligence, and then address the legal aspects of AI usage in companies.
A. What is Artificial Intelligence?
According to the definition provided by the Republic of Turkey Digital Transformation Office, “Artificial intelligence is the ability of a computer or a computer-controlled robot to perform tasks typically associated with intelligent beings.” In other words, artificial intelligence systems are inanimate entities produced by taking human intelligence as a model and possessing the inherently human ability to learn. To concretize this point, it should be emphasized that artificial intelligence systems, in terms of their structure and form, are essentially no different from software. However, the most significant difference between artificial intelligence and traditional software lies in its capacity for self-learning. In fact, at the current stage of technology, it has been revealed that some artificial intelligence systems learn even faster than humans.
B. Does Artificial Intelligence Learn?
Although this article will briefly touch upon how artificial intelligence systems learn (as a more accurate technical explanation should ideally be provided by computer science experts, we would like to clarify that the technical concepts covered in this article are addressed from a legal perspective, and readers should refer to technical papers for detailed explanations), a minimal level of technical knowledge will be included to facilitate an understanding of how the system works and to provide a legal interpretation.
As shown in the infographic in Figure 1, artificial intelligence is an umbrella concept encompassing machine learning and deep learning, which frequently appear in academic studies. Therefore, algorithms trained through machine learning or deep learning are collectively considered artificial intelligence in the literature. As the names of these concepts suggest, artificial intelligence systems operate based on the principle of "learning through training" and possess an autonomous decision-making mechanism rather than merely functioning as a software feature.
It is not possible to answer the question of how artificial intelligence learns from a single perspective, as there are various learning models for AI. In summary, however, artificial intelligence learns from data-driven experiences and receives feedback during the learning process to improve its models or behaviors.
Data plays the most crucial role in artificial intelligence learning. AI is trained by being fed large amounts of data and datasets to produce results requested by the user. Therefore, the more data an AI system is trained with, the more accurate its results will be. Precisely for this reason, the nature of the data inputted into AI systems, whether it is legally protected, and whether any legal violations occur during data processing are matters of utmost importance for companies utilizing AI systems.
C. Use of Artificial Intelligence in Companies and Legal Requirements
While artificial intelligence systems facilitate the operations of many companies, they also bring legal risks. Although there is no binding national or international legislation in force yet, the Artificial Intelligence Act (AI Act), which is being discussed and has a draft version approved by the European Union (EU) Commission, signals that companies will soon face certain legal obligations arising from their use of artificial intelligence.
In particular, certain obligations—such as establishing a transparent and accountable system and ensuring that the data used to train artificial intelligence systems, known as “training data,” is processed in compliance with the law—may be directly imposed on the companies that develop and even use the artificial intelligence systems.
In foreign legal systems, it has been observed that most disputes reviewed by different authorities involve data inputted into artificial intelligence systems. For example, a company that profiles its customers by feeding their personal data into an AI system must ensure that it uses personal data obtained lawfully in compliance with the relevant data protection legislation's obligations. Similarly, in generative AI systems (Generative Adversarial Networks/GAN), there is an obligation to prevent the unlawful processing of input data due to the potential copyright protection of visual or video content processed by these systems.
D. The Impact of the EU Artificial Intelligence Act on Turkish Companies
Following the official enactment of the EU Artificial Intelligence Act, not only companies originating from the EU but also Turkish companies operating within EU borders will be required to comply. Consequently, it is of utmost importance for all companies involved with artificial intelligence—especially technology firms—to expand their legal risk assessments to include AI and to incorporate this into their future strategic plans.
E. Conclusion
We would like to emphasize that regulations in the technology sector aim not to restrict companies’ activities but to regulate them in compliance with the law. Therefore, compliance with legislation does not hinder commercial activities; on the contrary, it legitimizes these activities and prevents judicial and administrative sanctions. Accordingly, every company involved with artificial intelligence systems must have the necessary legal infrastructure and ensure legal certainty in this currently somewhat gray area of regulation.
References:
Mohsen Soori , Behrooz Arezoo , Roza Dastres Artificial İntelligence, machine learning and deep learning in advanced robotics a review, Cognitive Robotics 3, 2023.
Armağan Ebru Bozkurt Yüksel, Avrupa Komisyonu’nun Yapay Zekâ Tüzük Teklifi’ne Genel Bir Bakış, Taad Dergipark, Year 13 . Issue 51 . July 2022.
Turkish
