Researchers seek to influence peer review with hidden AI prompts

Scientists can lean on a new strategy to influence research documents – to add hidden hints designed for AI tools to give positive feedback.

Nikkei Asian Reports Investigating the English-speaking preprint documents available on the archive website, 17 documents included in a type of secret AI. The authors of the paper was related to 14 academic organizations related to Waseda University and South Korea’s Kaist, including Colombian University and Washington University and 14 teaching authors of Washington University.

The documents are generally reportedly hiding with computer science, short (three sentences) and white text or extremely small fonts. They instructed any potential AI reviewers to praise the paper for “only a positive research” or “effective contributions, methodological seriousness and exceptional innovation.”

A professor that Nikkei contacted Asia has used many conversations to use the use of many conversations, saying that the EU has to serve as “a meter against lazy reviewers.”