Artificial intelligence tools can now mass-produce academic finance papers that are nearly indistinguishable from human-authored research, according to a study published in the Journal of Economic Literature. The findings raise questions about the integrity of academic publishing and peer review.
The research demonstrates that large language models can generate entire papers—complete with data analysis, citations, and formatting—that fool human reviewers. This capability threatens to overwhelm journals with low-quality or fraudulent submissions, potentially undermining trust in financial scholarship.
Study authors tested multiple LLMs and found their outputs matched human writing quality across grammar, structure, and logical flow. The models performed particularly well on standard finance topics where conventions are well-established.
The implications extend beyond academia. If financial institutions rely on such papers for investment strategies or regulatory compliance, the risk of acting on fabricated research grows. Journals may need to adopt AI-detection tools and stricter verification processes.
Experts caution that while LLMs can mimic form, they lack genuine understanding of economic principles. The study itself notes that AI-generated work may contain subtle but critical errors in reasoning.