The AI 'Ghosts' Contaminating Academic Publishing
Hacker News
Read full postResearchers from Samsung and the University of Warsaw identified recurring fictitious author names generated by large language models (LLMs) in AI-created academic papers and other documents. These AI-generated names, such as Elena Vasquez and Marcus Chen, appear repeatedly as co-authors despite not existing in reality. The study highlights how LLMs produce correlated name patterns, leading to the presence of 'ghost' authors contaminating academic publishing and online information.



