The emotional impact of generative AI: negative emotions and perception of threat

Abstract

ABSTRACT Generative Artificial Intelligence (AI) is a rapidly expanding field that aims to develop machines capable of performing tasks that were previously considered unique to humans, such as learning, reasoning, problem-solving, and decision-making. The recent release of several tools based on AI (e.g. ChatGPT) has sparked debates on the potential of this technology and garnered widespread attention in the mainstream media. Using a socio-psychological approach, in three studies (total N = 410), we demonstrate that when faced with Generative AI’s ability to reproduce the complexity of human cognitive capabilities, participants reported significantly higher negative emotions than those in the control group. In turn, negative emotions elicited by a specific type of AI (e.g. generative AI) were associated to the perception of threat extended to AI technologies as a whole, understood as a threat to various aspects of human life, including jobs, resources, identity, uniqueness, and value. Our findings emphasise the importance of considering emotional and societal impacts when developing and deploying advanced AI technologies and implementing responsible guidelines to minimise adverse effects. As AI technology advances, addressing public concerns and regulating its usage is crucial for the benefit of society. To achieve this goal, collaboration between experts, policymakers, and the public is necessary.