Regulation of artificial intelligence

The speed of the development of artificial intelligence systems and their impact on society has created the need for artificial intelligence to be regulated. How and in what way artificial intelligence will be regulated has been the topic of various discussions within the literature. In this thesis, the definition and process of the development of artificial intelligence is examined and the discussions regarding the regulation of artificial intelligence are mentioned. Then, regulations that are developed in order to regulate artificial intelligence, such as the Draft Artificial Intelligence Act and the Artificial Intelligence Liability Directive Proposal, were examined and the positive and negative aspects of these regulations were analyzed. Current studies within the Turkish doctrine and suggestions regarding to possible artificial intelligence regulations that could be developed within the Turkish legal system were also given. Afterwards, the concept of algorithmic regulation is defined and how artificial intelligence can be regulated with algorithmic regulation is explained through the idea that certain legal norms in the aforementioned regulations can be transformed into their digital representations and integrated into artificial intelligence systems as an algorithm. Then, it is stated how the relevant regulation and regulation methods could be applied to computer vision systems, which is a subfield of artificial intelligence, within the context of biometric systems. Lastly in the conclusion, suggestions are made regarding the future of algorithmic regulations and its development within the context of how the legal norms within the relevant regulations could be transformed to their algorithmic counterparts and be integrated to artificial intelligence systems.