Among 97·1 million verified references [across 2.5 million papers], we identified 4046 fabricated references across 2810 papers (illustrative examples are shown in the appendix p 5–6). In 2023, approximately one in 2828 papers contained at least one fabricated reference. By 2025, this had risen to one in 458 and in the first 7 weeks of 2026, one in 277 papers had at least one fabricated reference. The fabrication rate increased more than 12 times, from approximately four per 10 000 papers in 2023, to 51·3 per 10 000 papers in the fourth quarter of 2025, reaching 56·9 per 10 000 papers in early 2026 (figure).
Fake citations in philosophy papers won’t get people killed; these might. Science journals will need to adopt draconian rules (e.g., if you’re caught submitting a paper with AI-generated citations, you’re banned for ten years from submitting and, depending on defamation law in the jurisdiction, your wrongdoing will be publicized).
“First, you say just now: “One impression I may have given with my piece is that I think all would…