Talking AI and Talking Money
Author:
Zoey Zeng ’27Co-Authors:
Ethan CowenFaculty Mentor(s):
Kate Suslava, Accounting & Financial ManagementFunding Source:
Program for Undergraduate ResearchAbstract
From your homework to stock picks, AI search engines can do it all. In Q3 of 2025 alone, the proportion of S&P 500 firms discussing AI during earnings calls was 62%! Regardless of industry, AI will be affecting how a business operates and functions. At the same time, managers are grappling with different views on AI, ranging from optimism to concern. In this project, we analyze corporate disclosures by reading extractions of AI discussions and classifying these extractions into costs and benefits of AI. In our findings, positive AI-related keywords tend to emphasize growth and potential—such as opportunities, expansion, and innovation—whereas negative sentiment toward AI is more often associated with words like bias, concerns, and infringement. Next, we generate textual rules that extract costs and benefits of AI from 10-Ks, 10-Qs and earnings calls and calculate AI sentiment scores for each firm. Finally, we examine the trends of the AI sentiment across time and different sectors.