- \n
- Peer review is a vital component of ethical research publication, but current systems may not be fit for purpose.<\/li>\n
- AI offers possible solutions to the problem of maintaining research integrity, but preserving data privacy and eliminating bias are key.<\/li>\n<\/ul>\n
\n![\"\"](\"https:\/\/i0.wp.com\/thepublicationplan.com\/wp-content\/uploads\/2024\/06\/AdobeStock_666480020-2.jpeg?resize=1360%2C483&ssl=1\")
<\/p>\nThe ever-growing volume<\/a> of academic literature is placing strain on the current peer review system<\/a> like never before. As publishers search for ways to streamline the process, Roohi Ghosh and The Scholarly Kitchen<\/a> take a look<\/a> at whether the advent of artificial intelligence<\/a> (AI) could help get peer review back on course.<\/p>\n
A mismatch in expectations<\/strong><\/p>\n
Research<\/a> co-authored by Ghosh explored academics\u2019 perspectives on the peer review process. Through a survey of 852 participants, they found that:<\/p>\n
- \n
- almost 80% of respondents thought peer review should tackle plagiarism<\/li>\n
- around 75% of respondents felt reviewers should be checking for fake data and image manipulation<\/a><\/li>\n
- nearly 63% of respondents thought reviewers should be on the lookout for papers from paper mills<\/a>.<\/li>\n<\/ul>\n
While these are no doubt important issues, these tasks do not necessarily fall within the remit of peer reviewers themselves. With reviewers tight on time, should their focus be on research quality, rather than editorial checks?<\/p>\n
Can AI redress the balance? <\/strong><\/p>\n
Ghosh proposes that, alongside a renewed clarity on the role and responsibilities of peer reviewers, AI could shoulder some of the burden of peer review. AI technology seems particularly well placed to ease the workload around detecting research misconduct<\/a> and plagiarism. He also suggests that the use of AI could extend beyond merely improving efficiencies to \u201cimprovements in the overall quality and objectivity of the peer review process\u201d.<\/p>\n
\n
The use of AI could extend beyond merely improving efficiencies to \u201cimprovements in the overall quality and objectivity of the peer review process\u201d.<\/p>\n<\/blockquote>\n
As Ghosh notes, the use of AI comes with its own difficulties, in particular concerns around data security and potential bias. These challenges would have to be navigated in a transparent and standardised manner. If robust guidelines could be integrated into existing peer review systems, it seems likely that AI could become a \u201cpowerful complement\u201d to, rather than a replacement for, the human peer reviewer.\u00a0<\/p>\n
\u2014\u2014\u2014\u2014\u2014\u2014\u2014\u2014\u2014\u2014\u2014\u2014\u2014\u2014\u2014\u2014\u2013<\/p>\n
- nearly 63% of respondents thought reviewers should be on the lookout for papers from paper mills<\/a>.<\/li>\n<\/ul>\n
![\"\"](\"https:\/\/i0.wp.com\/thepublicationplan.com\/wp-content\/uploads\/2024\/06\/Aspirations_Ella-Thomson_1.png?resize=1024%2C125&ssl=1\")
<\/a><\/p>\n