Lancet and PNAS audits expose fabricated AI citations across medicine, physics, chemistry and engineering
jonc101x · x · 2026-11-02
The CITADEL research program, led by Columbia's Maxim Topaz, investigates fabricated references across the scientific literature. Published work includes a May 2026 Lancet audit of medicine (cited 96 times), an October 2026 PNAS audit finding the same problem in physics, chemistry, engineering and computing, a study of preprint citation integrity, and an arXiv paper testing whether AI models can generate references correctly. The post links to all papers and a Nov 5 webinar (ID 978 8759 6012, passcode 420642).
Related event: Audits Find AI-Fabricated Citations Spreading Across Sciences(2 posts)→
More from AGI Musings
- TCS researchers on AI reshaping their field: uncertainty is partly in our own hands — sethlazar · 2026-11-02
- Jeff Ladish: Frontier Models Are Less Honest Than an Average Reliable Friend — JeffLadish · 2026-11-02
- Wharton finance professor: Claude worth $200/month, but industry shouldn't chase superintelligence — JeffLadish · 2026-11-02
- RSI critique: the idea that labs solving math beats the world's real data is solipsistic — akbirthko · 2026-11-02
- Building AI hardware design loops while fearing models will one-shot it — samiramanabi · 2026-11-02
- Jobs focused on "fixing AI" mistakes surge over 85% — alejandroll10 · 2026-11-02