Using AI and Behavioral Finance to Cope with Limited Attention and Reduce Overdraft Fees.

This article examines the effectiveness of an artificial intelligence (AI)–based human–algorithm interaction designed to reduce overdraft and nonsufficient fund (NSF) fees among users of Mint, a large personal financial management platform in the United States and Canada. Through a randomized field experiment involving 39,607 users, the study finds that sending as-needed email reminders significantly reduces overdraft fees, with simpler and negatively framed messages proving more effective than more complex or positively framed ones. The intervention was most effective for users with medium to high incomes and fair to good credit scores, while those facing greater financial hardship showed less responsiveness, indicating that AI reminders alone may not suffice to overcome financial constraints. The research contributes to understanding AI-based financial advising, human–computer interaction, and behavioral finance by highlighting the role of message design and user characteristics in influencing financial behavior.