Cryptocurrency Investment Decisions in the Age of Artificial Intelligence: Behavioral Biases and Algorithmic Tools

Abstract

The emergence of cryptocurrencies has transformed individual investment practices, exposing millions of investors to highly volatile markets where behavioral biases significantly influence decision-making. Simultaneously, artificial intelligence tools (robo-advisors, trading algorithms, and predictive analytics) have become increasingly accessible to retail investors, promising decisions that are more rational. However, the interaction between these two phenomena remains largely unexplored. This conceptual paper addresses the question: how does the use of AI tools moderate the influence of traditional behavioral biases on cryptocurrency investment decisions? Drawing on a systematic literature review at the intersection of behavioral finance, cryptocurrency studies, and human-computer interaction, we develop an integrative conceptual model. The model incorporates four cognitive biases (representativeness, overconfidence, herding behavior, loss aversion) as independent variables, cryptocurrency investment decision as the dependent variable, and AI tool usage as a moderating variable. We propose that AI exerts a contrasted moderating effect: mitigating certain biases (representativeness, herding, and loss aversion) while potentially amplifying others (overconfidence) and creating new vulnerabilities such as automation bias. Seven research propositions are formulated for future empirical testing. This conceptual framework contributes to the literature on behavioral finance in digital contexts and offers practical implications for investors, platform designers, and regulators concerned with responsible innovation and investor protection in the age of AI.