Etik inom AI - en idé eller verklighet? : En studie om IT-konsulters arbete med etik

Sammanfattning

The development of Artificial Intelligence has increased massively in recent years. Companies have been restructured and work processes have been streamlined. Even though many don't think about it, the phenomenon now exists in most contexts, even in highly sensitive activities. But despite the many possibilities with AI, there are also a lot of ethical challenges that need to be managed. The research in this topic focuses on identifying ethical challenges with AI, rather than understanding how they are handled practically. Therefore, the purpose of this study is to investigate how IT-consults work with AI:s ethical challenges during the development and implementation of AI-solutions. The method for collecting data was qualitative and was carried out according to a three-step model consisting of informal and semi-structured interviews with IT consultants and AI experts. A total of four respondents participated in the study, and to analyze the empirical material, an inductive thematic analysis was used. The results show that the handling of AI's ethical challenges took place heterogeneously through the work process. Although the respondents showed a good understanding of the ethical aspects, these were given low priority in relation to the customer's needs and business value. In addition, the results showed that the responsibility between the different actors in AI projects was distributed unevenly. The study concludes that the work with ethical aspects in the development and implementation of AI solutions is situational and places high demands on the individual rather than the organization. The study contributes to research by providing a practical perspective on how ethical challenges are handled in AI projects and highlights the need for, for example, clearer internal ethical guidelines and increased transparency from suppliers.