Artificial intelligence is shaking up the academic world, raising serious concerns about its role in the peer review process.
The Growing Issue of AI-Generated Peer Reviews
The scientific community is grappling with an emerging challenge: how to handle AI-generated peer reviews. This issue came to light when ecologist Timothée Poisot received feedback that was unmistakably produced by ChatGPT. The review contained a revealing phrase: “Here is a revised version of your review with improved clarity and structure.” Poisot was outraged, arguing that such AI-generated responses undermine the core principle of peer review—expert evaluation by human scholars.
“I submit a manuscript for review in the hope of receiving insights from my peers,” Poisot stated in a blog post. “When that expectation is not met, the entire foundation of peer review collapses.”
AI’s Growing Influence in Academic Research
Poisot’s case is not an isolated incident. A study published in Nature revealed that up to 17% of AI conference paper reviews from 2023-24 exhibited substantial modifications using language models. Moreover, a separate survey found that nearly 20% of researchers admitted to leveraging AI tools to expedite the peer review process.
This has led to bizarre consequences. In 2024, a paper published in the Frontiers journal contained nonsensical diagrams generated by the AI art tool Midjourney. One figure even depicted a distorted rat, while others featured indecipherable symbols and gibberish text. The fact that such flawed content made it through peer review alarmed many academics.
Publishers Respond with Stricter Guidelines
In response to these issues, major publishers have begun implementing stricter policies:
- Elsevier has banned the use of generative AI in peer review entirely.
- Springer Nature and Wiley permit limited use of AI tools, provided their application is disclosed.
- The American Institute of Physics is cautiously experimenting with AI tools to supplement—rather than replace—human review.
The Debate: Can AI Enhance Peer Review?
Despite concerns, some researchers believe AI can be beneficial if used carefully. A Stanford study found that 40% of scientists considered ChatGPT-generated reviews to be just as helpful as human reviews, while 20% found them even more useful.
However, critics remain skeptical. Poisot and other academics argue that peer review should be based on human expertise, not an algorithm’s automated response. “If we don’t push back against AI-generated reviews, we are conceding defeat,” Poisot warned.
The Future of AI in Scientific Publishing
The debate over AI’s role in peer review is far from settled. Some believe that AI, when used responsibly, could streamline the process and improve efficiency. Others fear that unchecked AI usage could erode the credibility of academic publishing.
As publishers tighten regulations, researchers will need to decide how much they are willing to rely on AI in their work. One thing is certain—the future of AI in academic research will continue to be a contentious issue.
For further exploration of AI advancements in research and technology, check out Gemini 2.5: Google Unveils Its Most Advanced AI Model Yet.
