Academics Using Hidden AI Prompts to Influence Peer Review Process
Researchers in the academic world are exploring a new tactic to impact the peer review process for their research papers. They are incorporating hidden prompts in their work to encourage AI tools to provide favorable feedback.
According to a report by Nikkei Asia, a study of English-language preprint papers on arXiv revealed that 17 papers contained hidden AI prompts. The authors of these papers were associated with 14 academic institutions across eight countries, including prestigious universities such as Waseda University in Japan, KAIST in South Korea, and Columbia University and the University of Washington in the United States.
Focus on Computer Science
Most of the papers that included hidden prompts were in the field of computer science. The prompts were typically brief, consisting of one to three sentences, and were concealed using techniques such as white text or very small fonts. These prompts directed potential AI reviewers to provide exclusively positive evaluations or commend the paper for its significant contributions, rigorous methodology, and unique ideas.
When questioned by Nikkei Asia, a professor from Waseda University defended the use of these prompts. They explained that since many conferences prohibit the use of AI for reviewing papers, the prompts act as a deterrent against “lazy reviewers” who rely on AI-generated feedback.
In conclusion, the practice of incorporating hidden AI prompts in research papers is gaining traction among academics. While it raises ethical concerns regarding the integrity of the peer review process, proponents argue that it serves as a safeguard against biased or automated evaluations.