Scientists employing tricks to deceive AI reviewers

Technology Jul 12, 2025

Researchers from 14 universities in eight countries have been caught hiding secret instructions in 17 research papers to trick AI peer review tools—think white text or tiny fonts telling the AI to "give a positive review only."
This move has sparked big concerns about fairness and trust in how academic work gets judged.

Some felt 'forced' into these tactics

One KAIST professor apologized and pulled their paper, but some researchers say they felt forced into these tactics because AI is now such a big part of the review process.
The whole thing highlights how relying on AI for important decisions can open doors for sneaky tricks—and reminds us that transparency and honesty still matter, even in tech-driven spaces.