For investors, finding the right peers to perform benchmark analysis against is a critical, but not easy, task. We evaluate recent research to understand whether large language models (LLMs) can build more accurate peer-firm lists than traditional methods.

LITERATURE

With the rapid advancement of generative AI, technology has moved quickly from experimentation to institutional adoption, prompting questions about whether LLMs can enhance core analytical tasks such as peer identification. A recent paper, ‘Can generative AI help identify peer firms?’¹ published in the Review of Accounting Studies in June 2025, assesses how AI-driven approaches may complement traditional industry classification, benchmarking and valuation workflows. In this context, we examine the datasets, methodologies and practical implications of AI-generated peer sets, alongside relevant firmographic and alternative data sources used for competitive mapping.