GPTZero Analysis Finds 100+ Hallucinations in NeurIPS 2025 Accepted Papers
By
segmenta
A five-star bake. Worth schmearing, sharing, saving.
Summary
GPTZero's analysis of 4,841 papers accepted by NeurIPS 2025 reveals at least 100 papers contain confirmed hallucinations, including fabricated authors and mismatched citations. The AI detection tool identified instances where papers referenced non-existent authors or linked to different articles than cited, raising concerns about academic integrity in AI research publications.
Key quotes
· 4 pulledGPTZero's analysis of 4,841 papers accepted by NeurIPS 2025 show there are at least 100 with confirmed hallucinations
Authors are obviously fabricated. arXiv ID links to a different article
Article with a matching title exists here. Authors are obviously fabricated
Unmasking Puppeteers: Leveraging Biometric Leakage to Expose Impersonation in AI-Based Videoconferencing
You might also wanna read
Columbia researcher's own paper nearly published with AI-hallucinated reference, highlighting growing academic integrity concern
A Columbia University associate professor, Maxim Topaz, who researches AI applications in healthcare, nearly published a scientific paper co
fortune.com·4d ago
Columbia professor studying AI nearly publishes paper with AI-hallucinated reference, highlighting growing academic integrity concern
A Columbia University associate professor, Maxim Topaz, who researches AI applications in healthcare, nearly published a scientific paper co
exacted.me·4d ago
Study Finds Over 2,800 Biomedical Articles Contained AI-Fabricated Citations
A research correspondence published in The Lancet reveals that over a three-year period, 4,046 references in 2,810 published scientific jour
forbes.com·1d ago
AI-generated research papers overwhelm academic peer review and citation systems
The article discusses a growing crisis in academic publishing where AI-generated research papers are flooding journals and citation database
The Verge·16d ago
OpenAI says GPT-5.5 Instant reduces ChatGPT hallucinations by over 50% on high-stakes prompts
OpenAI claims its new GPT-5.5 Instant model, now the default for ChatGPT, hallucinates significantly less than the previous GPT-5.3 Instant
The Verge·26d ago
Fake academic journals publish AI-generated papers under real professors' names
A network of fake academic journals has published over 100 AI-generated papers attributed to real professors without their knowledge. The fr
nbcnews.com·4d ago