AI-hallucinated citations are creeping into papers that shape clinical guidelines, researchers warn

Editor's take

Fabricated citations, a phenomenon previously seen in less critical contexts, are now appearing in biomedical research papers that influence clinical guidelines, with rates escalating dramatically to over twelve times their previous levels since 2023. This trend is particularly concerning because these papers inform medical decision-making, potentially impacting patient care and public health if based on non-existent evidence. The issue highlights the growing challenge of maintaining scientific integrity as generative AI tools become more prevalent in academic workflows.

The rapid increase suggests a direct correlation with the adoption of large language models, such as those developed by OpenAI or Google, in academic writing and literature review. This development necessitates urgent scrutiny of AI-generated content within the scientific community, particularly in fields with direct real-world consequences. Future research should focus on developing robust detection mechanisms and establishing clear guidelines for AI use in scholarly publishing to prevent the propagation of misinformation.