Recent findings reveal that artificial intelligence is creating fake references to non-existent medical research, posing a significant threat to the integrity of clinical guidelines. An audit of millions of biomedical papers uncovered over 4,000 citations to fabricated studies, as reported in The Lancet. These false citations can mislead healthcare professionals, potentially leading to treatment decisions based on non-existent research. Maxim Topaz, an associate professor at the Columbia School of Nursing and the study’s lead author, highlighted the growing prevalence of such errors, which have increased twelvefold in the past three years. Topaz’s investigation was prompted by his own experience with an AI tool that inserted a fake citation into his work, which nearly went unnoticed through peer review. The issue arises when AI tools generate citations for statements of fact, sometimes attributing them to real authors or entirely fabricating them. This problem is not limited to medical research and could affect other fields as well. The realistic appearance of these AI-generated citations underscores the need for rigorous fact-checking by researchers.
QUESTION: How might the increasing presence of AI-generated fake citations impact the trustworthiness of scientific research in the future?
AI is fabricating citations in biomedical studies, researchers find
