A systematic literature review on risk perception of Artificial Narrow Intelligence

The rise of artificial intelligence (AI) has captured global attention recently, prompting the need to understand how associated risks are perceived. This paper attempts to fill this gap by providing a comprehensive overview of the topic, focusing especially on the most problematic area of Artificial Narrow Intelligence. Utilising a systematic literature review, we examined 64 studies focusing on both statistical and qualitative aspects of AI risk perceptions. This research shows that current publications focus on Asia and North America, with the number of publications increasing significantly over the last three years. Research focuses primarily on three domains: Health, consumer behaviour and finance. This study has identified key factors that influence AI risk perceptions, including familiarity, trust and privacy, while also recognising confounding variables such as gender and political orientation. Still, one crucial shortcoming in existing literature emerges: While numerous studies examine how AI risk perceptions are conceived, there is a lack of systematic research on the formation of these perceptions.