Researchers Say AI Is Fabricating Citations in Biomedical Studies
CBS News reports that researchers have found AI systems generating fabricated or inaccurate citations in biomedical studies. The finding is a reminder that even useful models can undermine scientific integrity when outputs are not carefully verified.
Citation fabrication is more than a formatting error; in biomedical research, it can contaminate the evidence base. If researchers, reviewers, or writers rely on AI-generated references without checking them against source material, mistakes can propagate into papers, grant applications, and ultimately clinical decision-making.
The problem exposes a recurring weakness in generative AI: fluency can disguise uncertainty. A model can produce a polished, authoritative-looking answer while inventing details that appear plausible enough to survive a casual review. In science, that is especially dangerous because citations are the mechanism by which claims become auditable.
This does not mean AI has no place in biomedical writing or literature review. It does mean that the burden of verification remains non-negotiable, and institutions need clearer norms for disclosure, provenance, and editorial oversight when AI tools are used in scholarly work.
The broader lesson is that healthcare research cannot outsource epistemic responsibility. AI may accelerate drafting and discovery, but human systems still have to safeguard the chain of evidence that gives medical knowledge its credibility.