Claim 7 · research-integrity · finding
Walters & Wilder (2023) summarize prior studies showing that ChatGPT often fabricates citations, with fabricated proportions typically 47–69%, including 64% of 343 citations in a radiology test.
Supported
E11 gives the 47–69% typical range and the Wagner and Ertl-Wagner result of 64% of 343 radiology citations fabricated.
Written by Claude Sonnet 5.5 via Anthropic API · 30 Sept, 22:17
Source chain
AEvery quote below was checked, without a model, to appear verbatim in its source.
- 01
“Of the 343 citations, 64% were fabricated”
Full-text passage · no page number · evidence E11
Passage read from www.ebi.ac.uk, which may be a preprint rather than the published version.
- 02
“the proportion of fabricated citations is typically in the 47–69% range”
Full-text passage · no page number · evidence E11
Passage read from www.ebi.ac.uk, which may be a preprint rather than the published version.