The advent of artificial intelligence (AI) has shaken almost every industry across the world—and the scholarly publishing industry is no exception. With the rapid advancement of AI and its sudden availability to the public, many find excitement in the potential it has to automate processes such as peer review, and many are concerned about the implications that this technology has when implemented into these processes. In a world where more and more papers are sent for publication every day with peer reviewers dwindling, many see AI as a potential to automate these processes to lessen the load on experts volunteering their time to review these publications. However, others see it as a risk to confidentiality and academic integrity.
Benefits
As many that have experimented with AI tools would be able to testify, AI tools do a good job of correcting and cleaning up language in writing in a smooth and coherent way. This is especially useful for international peer review, where non-native language speakers may use these tools in articulating their notes clearly. Overall, this enhances global scholarly communication and provides opportunities for experts who otherwise may not have participated in peer review.
Additionally, while AI tools are still in their infancy, scholars have already started experimenting with their use in automating peer review tasks. It was found that when focusing on superficial features like word distribution, readability and formatting scores, machine learning systems would often be able to predict peer review outcomes (Checco, 2021). With this, some could propose automated screening to assist in picking out papers that may be rejected anyway, saving reviewers valuable time. Additionally, automated screening tools could possibly be used to address limitations of peer review such as bias, spotting ambiguous/incorrect citations, missing information, incorrect calculations, etcetera (Schulz, 2022).
Because these are based on more superficial features, the accuracy of automated AI tools in spotting these problems could very well be because they are features that negatively impact reviewer’s “first impression bias”—if the readability is poor, both the reviewer and AI may both advocate a rejection despite the research being sound. With this in mind, while the tool may not be able to help automate the peer review process properly, the researchers could still use it to screen their papers prior to review and receive advice to rework the papers to eliminate this bias (Checco, 2021).
Challenges
One of the main concerns with AI and automation in peer review is simply that it lacks the critical thinking skills that a human can provide. While AI can mimic ideas, piece information together, and even assist with writing at the grammatical level, AI does not hold the expertise that a human reviewer can provide. Especially with the gap between up-to-date information that is available to a human reviewer versus the delay in data that an AI platform can have access to. Additionally, when met with a gap in knowledge, AI tends to make up information to fill said gaps, making it unreliable in fully automating any processes in academia.
Of course, many are not advocating for AI to wholly take over peer review. Even as a tool to assist reviewers in automated processes such as screening, it is essential to keep these limitations in mind. Much like humans, AI is fallible, and it is very much capable of making mistakes. Tools assessing quality may not capture information that reflect methodological quality and cannot determine whether methods used are suitable to answer research questions (Schulz, 2022).
Beyond all of this, many are hesitant about the implications of using these tools during the review process because of the confidentiality risks associated with them. Uploading a manuscript to these tools can provide them with sensitive or confidential information. The use of certain AI tools is giving information directly to a third party that does not necessarily treat sensitive information as confidential.
Where We Stand
Given the challenges and concerns in using these tools in peer review, IGI Global’s AI policy states that because manuscripts under peer review may contain sensitive or confidential information, they should not be shared outside the peer review process—including any generative AI tool or service. This would be considered a breach of confidentiality and privacy. IGI Global does not permit editors and peer reviewers to upload an unpublished manuscript or any information pertaining to the manuscript into Generative AI tools.
Overall, it is the peer reviewer’s responsibility to ensure the accuracy and integrity of the research and to formulate their own opinions and recommendations. Allowing AI to assist with decision making, using it to vet accuracy and integrity, etc. are not permitted.
If peer reviewers require AI tools to support with the grammar and/or improve the readability of the evaluation that they have written, they may do so, but they must disclose the AI tool and its use within the peer review.
The policy is very standard and can be seen reflected in other publishers’ policies as well.
Looking Ahead
AI is a new technology that will not be going away anytime soon. As it develops, it will be harder and harder to ignore its benefits in process automation. What is almost guaranteed is that the debate will continue on the ethics of AI and automation’s use in publishing and peer review. As technology continues to develop, it is possible that some of these challenges will be addressed, and these tools can be harnessed.
Should this be the case, many are in agreement that AI and automation should never take over peer review entirely, but should be used in conjunction with real, human experts.
IGI Global is always looking to advance the discussion of AI in not only the peer review process, but also in academia and beyond. Researchers looking to contribute their expertise can browse IGI Global books and journals seeking chapters and articles on these and more topics on the
Publications Seeking Submissions page. For those looking to edit or author a full book, proposals may be submitted
here.
About IGI Global – Publishing Tomorrow’s Research Today
Founded in 1988 and headquartered in Hershey, Pennsylvania, USA with a subsidiary office (IGI Science and Technology, Ltd.) in Beijing, China, IGI Global is a rapidly growing, mid-sized, independent international academic publisher dedicated to releasing high-quality, peer-reviewed scholarly research publications in the three major academic subject areas of
Business & Management,
Science, Technology, & Medicine (STM), and
Education & Social Science. Our commitment is to advance the discovery of pioneering scientific research promptly through swift and agile processes, with books released in approximately 6-9 months from start to finish. With more than 200,000 research collaborations with experts from leading institutions around the world, IGI Global anticipates publishing over 1,000 books during 2025 copyright year, positioning our publishing output among the industry’s giants with centuries of history. Our Open Access Journal Program includes nearly 200 fully open access journals, making IGI Global one of the largest 100% open access journal publishers in the world. Through traditional book publishing and open access options, IGI Global strives to bring cutting-edge high-quality research at the pace of innovation, ensuring that
tomorrow’s research is published today.
Learn more about IGI Global
here.