最新报道:Researchers from Stanford and Princeton have developed a method to embed hidden AI prompts in academic preprints that subtly influence peer reviewers. The technique, called "suggestive prompting," uses steganography to conceal AI-generated suggestions within manuscript text that nudge reviewers toward specific conclusions without their awareness. Early tests show these prompts can increase acceptance rates by up to 30% by guiding reviewers' attention to study strengths. While proponents argue it could reduce bias, critics warn it constitutes unethical manipulation of the peer-review process. The team has open-sourced their code but faces backlash from journal editors calling for disclosure requirements.