Abstract
<jats:p>Using AI in the form of a self-learning algorithm that can solve recurring problems based on available data, may lead to a development of a systemic deviation between the modeling and reality. Given deviation may result in bias in the decision-making process. This paper aims to highlight the phenomenon of AI-caused bias in insurance industry since insurers will rely on AI greatly. Since the bias has the potential to cause discrimination and unequal treatment of policyholders, the authors examine national legislation regarding the prohibition of discrimination in order to determine to what extent these regulations should apply to bias in insurance and what potential sanctions exist for such behavior. Additional issue to this matter is the question of liability for the AI-caused damage which required going beyond the national legislation and searching for the answers in newly adopted European legislation.</jats:p>