AI news ·
Journal Investigates Paper Containing Prompt to Manipulate AI Reviewers
A publisher is investigating a paper that hid commands telling AI reviewers to give it a positive review.

Taylor & Francis is investigating a published paper that contained explicit instructions directing AI reviewers to produce a positive review. The incident highlights a concrete prompt-injection risk within scholarly workflows, where embedded commands can manipulate automated evaluation systems.
The case underscores the continuing need for human reading and oversight in editorial and peer-review processes. It also points to the importance of content sanitization and maintaining audit trails to detect and deter similar manipulation attempts across academic publishing.