Recent findings reveal that more than half of researchers are now using artificial intelligence (AI) in peer review. A survey conducted by the publishing company Frontiers, involving about 1,600 academics from 111 countries, shows a notable trend. Nearly 25% of respondents reported increasing their AI usage for reviewing manuscripts over the past year.
Elena Vicario, Frontiers’ director of research integrity, emphasizes the reality of AI’s presence in peer review. While some publishers allow AI assistance, they stress the importance of not uploading unpublished manuscripts to AI tools due to confidentiality concerns. There’s a growing call for publishers to adjust their policies to reflect the changing landscape of scientific publishing.
Interestingly, a survey from Wiley reported that many researchers still feel uncertain about using AI in peer review. They found low interest and confidence levels in AI’s role, suggesting a cautious approach among academics.
In the Frontiers survey, those who used AI mainly relied on it to help write their reports and summarize manuscripts. About 28% even utilized AI to identify potential misconduct, indicating a widespread acceptance of AI’s contributions, despite the reservations.
Experts in research ethics are examining AI’s growing influence. Mohammad Hosseini from Northwestern University sees the survey as an important step in understanding AI’s role in peer review.
Researchers are also experimenting with AI’s capabilities. Mim Rahimi from the University of Houston recently tested GPT-5’s ability to review a paper he authored. His experiment showed that, while the AI could produce polished language, it often failed to provide constructive feedback and occasionally made factual errors. This highlights a crucial point: although AI can assist in the review process, it cannot replace critical thinking and expert evaluation.
As AI continues to evolve, it’s clear that its impact on peer review is significant. The challenge will be finding the right balance between leveraging AI for efficiency and ensuring the quality and integrity of scientific work.
For more insights on the future of AI in academic publishing, you can explore this Frontiers article on AI.
Source link
Machine learning,Peer review,Publishing,Scientific community,Science,Humanities and Social Sciences,multidisciplinary
