Weekly digest: Nelson memo block, OA books and artificial intelligence

Mark Elms

This week, we read the Fully OA group’s objection to a bill from the US House Committee on Appropriations that blocks the implementation of the OSTP open access mandate for federally funded research. We learn about the Direct to Open initiative achieving its second-year goal and about combining human intelligence with AI tools to root out unethical use of AI in research. We also read about how AI can be used to identify anonymized authors during peer review, about how scholarly publishing could use AI for its own benefit, and about improving preprint awareness in India. Finally, we watch a webinar from OASPA on the role of preprint peer review in scholarly communication.

To read:

Fully OA group respond to bill blocking open access via Fully OA | 3-minute read

The Fully OA group of open access publishers and open science platforms has written a response to the US House Committee on Appropriations’ bill on the Fiscal Year 2024. The response claims that the language within the bill “blocks implementation of the Office of Science and Technology Policy’s August 2022 guidance (the ‘Nelson memo’) to make federally funded research freely available without delay”. In their objection to the new bill, the Fully OA group argue that it is “detrimental” to the end goal of fully open science and will prevent American taxpayers from seeing the societal benefits of more than $90 billion worth of federally funded research. The Fully OA response was signed by leaders from eLife, Frontiers, JMIR Publications, MDPI, Open Library of Humanities, PeerJ, PLOS and Ubiquity Press.

Direct to Open initiative achieves second-year open access books goal via STM Publishing News | 8-minute read

The Direct to Open (D2O) initiative, created by the MIT Press in 2021, has now signed open access agreements with 322 libraries and consortia, allowing it to provide public open access to more than 160 books, scholarly monographs and edited collections. The D20 initiative has now hit its second-year goal, with 82 new books being made open access and available on the MIT Press Direct platform in 2023.

New approaches to combatting unethical AI use in research publications via The Scholarly Kitchen | 10-minute read

Tools that detect the use of artificial intelligence (AI) in writing are available, but simple workarounds – such as making minor changes to AI-generated text or using paraphrasing tools – can render them inadequate. This article by Minhaj Rais (Senior Manager of Strategy and Corporate Development at Cactus Communications) highlights the potential utility of combining human intelligence with AI-detection tools in the fight against the unethical use of AI in research publishing. For example, AI could be used to flag potentially fraudulent articles, which are then checked by a human subject matter expert before any binding editorial decisions are made.

AI can identify anonymized authors during double-blind peer review via London School of Economics | 5-minute read

For years, the double-blind review process has ensured fair and unbiased peer review by anonymizing both authors and reviewers. However, a recent research article has found that AI can be used to identify certain authors based on writing style, text content and cited references. With a data set of over 2 million articles and thousands of authors, the researchers found that authors with at least 12 publications were reliably identified by their AI model. This blog post explains the research in more detail and discusses the potential implications.

Can publishers use AI for their own benefit? via The Scholarly Kitchen | 10-minute read

While lots has been written about the potential problems with using AI in scholarly publishing, relatively little has been said about its potential benefits. This blog post, written by Avi Staiman (Founder and CEO at Academic Language Experts), discusses some of the ways that scholarly publishing can leverage AI for its own benefit, and what this might mean for open access publishing and for research authors.

The benefits and limitations of preprints in India via International Science Council | 6-minute read

A 2023 preprint reported that researchers from the USA and Europe were more familiar with preprints and had higher rates of preprint adoption compared to researchers from the rest of the world, including India. In response to the apparently low levels of preprint awareness among Indian researchers, the Indian National Young Academy of Sciences (INYAS), the International Science Council (ISC) and the Department of Science and Technology’s Centre for Policy Research conducted a joint online workshop to teach researchers about preprints and their potential benefits. This blog post from the ISC provides a summary of the workshop, which you can also watch in full on the INYAS YouTube channel.

To watch:

OASPA webinar on preprint peer review via OASPA | 1-hour watch

Catch up on an Open Access Scholarly Publishing Association (OASPA) webinar focusing on the role of preprint peer review. This webinar, which took place on 27 July 2023, features experts from the publishing, peer review and preprint fields, including Jonny Coates (Associate Director at ASAPbio), Lesley Anson (Steering Group Member at Science Colab), Stefano Bertozzi (Editor-in-Chief of Rapid Reviews: COVID-19), Thomas Guillemaud (Co-founder of Peer Community In), Sara Monaco (Managing Editor at Review Commons) and Daniela Saderi (Co-founder of PREreview). You can also watch the webinar on OASPA’s YouTube channel.

Enjoy reading our content? Then make sure to follow us on Twitter and LinkedIn for regular updates!