Generative AI in Capital Markets: Information Production, Dissemination, and Processing

• 成果类型: Article
• 署名作者: Cao, Sean Shun; Chen, Wilbur Xinyuan; Ma, Guang; Srinivasan, Suraj
• 署名单位: University System of Maryland; University of Maryland College Park; Hong Kong University of Science & Technology; Rutgers University System; Rutgers University New Brunswick; Harvard University
• 刊物名称: JOURNAL OF ACCOUNTING RESEARCH
• ISSN/ISSBN: 0021-8456; 1475-679X
• DOI: 10.1111/1475-679x.70061
• 发表日期: 2026-06
• 页码: 1427-1450
• 关键词: disclosure; earnings; cost; READABILITY; asymmetry

摘要

We synthesize evidence from six papers presented at the 2025 Journal of Accounting Research Conference on how generative artificial intelligence (GenAI) is reshaping capital-market information flows. Our discussion is organized around an economic framework with three layers: information production by firms and accounting professionals, information dissemination through intermediaries, and information processing by investors. Across these layers, the conference papers show that GenAI can lower preparation costs, improve intermediary productivity, and reduce investors' processing costs. At the same time, they point to a common constraint: whether GenAI improves the information environment depends critically on information verification costs. We also highlight gaps in current evidence and outline future research opportunities within and across the three layers.