Requirements for the Concept of the Implementation of Artificial Intelligence as a Tool for Preventing Corruption in the Public Sector

Abstract Artificial intelligence (AI) technologies have enormous potential for preventing corruption in the public sector. However, at present, the use of AI in the public sector is not regulated. Consequently, the article aims to propose concrete steps and actions for the implementation of AI technology in the public sector with a view to using this technology as a new tool to identify and prevent situations of corruption. The author conducted a study of Latvian state and local government officials in different regions and positions throughout 2023. In the article, the author analyses the research data on officials’ experiences with AI, their expectations about the effectiveness of AI in preventing corruption, as well as their basic requirements for the concept of implementation of AI. The author of the article proposes to create a basis for the legal environment for the use of AI and recommends adopting the guidelines of AI on the following basic ethical standards: the priority of human well-being; prohibition of harm at the initiative of the AI system; human control; compliance of design with law; prevention of covert manipulation of human behaviour; built-in security. In the article, the author recommends creating a special procedure for testing and further implementation of solutions of AI, bypassing redundant administrative procedures. In addition, the author recommends stimulating anti-corruption institutions and developing approaches to issues of civil liability for mistakes made by AI technology.