A new research paper from Samsung and the University of Warsaw highlights the phenomenon of AI-generated "ghost" authors appearing in academic publishing. These non-existent individuals, such as Elena Vasquez and Marcus Cheng Chye Tan, are repeatedly produced by large language models like ChatGPT, Gemini, and Claude when tasked with creating expert personas. The paper, "The Ghost Couple: Correlated LLM Name Priors and Their Haunting of the Web and Academic Publishing," demonstrates that these AI-generated names are not only consistent across different models but can also appear together as correlated ensembles. This contamination extends to academic repositories like Zenodo, where AI-authored papers with fabricated details are being indexed, potentially impacting the integrity of scholarly records and search aggregators like Google Scholar and ResearchGate. AI
IMPACT Raises concerns about the integrity of academic records and the potential for AI-generated content to be indistinguishable from human work.
RANK_REASON Research paper detailing a novel issue in AI-generated content impacting academic publishing.
Read on Mastodon — sigmoid.social →
- 404 Media
- Mastodon
- CERN
- ChatGPT
- Claude
- Elena Vasquez
- Gemini
- Google Scholar
- Marcus Cheng Chye Tan
- ResearchGate
- Research Gold
- Samsung
- The Ghost Couple: Correlated LLM Name Priors and Their Haunting of the Web and Academic Publishing
- University of Warsaw
- Zenodo
AI-generated summary · Google Gemini · from 3 sources. How we write summaries →
