While reading a chapter from a 2024 book on restoring biodiversity in post-industrial land earlier this month, an ecologist came across a reference to a study on rewilding she wanted to look up.
But the ecologist, who has asked to remain anonymous for fear of retaliation from the chapter authors, couldn’t find a record of the 2020 paper, purportedly published in Ambio by “Knijn et al.” She decided to check the other references. She told Retraction Watch she found at least eight that didn’t exist.
The ecologist flagged the issue to the book’s publisher, Wiley, on September 7 but told us she never heard back, aside from an automated response.
A Wiley spokesperson told us the publisher “takes integrity seriously and investigates all concerns when they are flagged to us.” They said Wiley would “take appropriate measures once the investigation has concluded.”
The publisher owned Edifix, a reference-checking service, from 2019 until it was acquired by another company earlier this year. The spokesperson did not answer our question asking whether books, including this one, were run through the program, but said Wiley “has published AI use guidelines for book authors and safeguards the integrity and quality of the books we publish through established checks at three stages: book proposal, contracting, and manuscript review and development. Edifix is not one of the tools that we currently use.”
The Wiley representative did not respond when we asked if the publisher was aware of issues with the book chapter before we contacted them last week.
Bhanu Pandey, one of the four authors of the chapter in the book, Restoring Biodiversity and Ecosystem Services on Post-Industrial Land, told us he wasn’t aware of the fake citations before we reached out. All four authors were researchers at CSIR-Central Institute of Mining and Fuel Research in Jharkhand, India, at the time of the study.
Bhanu Pandey told us first author, Sameer Shekhar, “took the lead in compiling the chapter materials and coordinating the submission.” He said he is looking into the book chapter and planning to submit a request for a correction to Wiley: “Where a citation is inaccurate or the associated statement is unsupported, I will propose the appropriate correction, revision, or removal and submit it to the chapter editor and publisher through their formal editorial process.”
None of the other authors responded to our request for comment, including Shekhar, who we reached out to multiple times on both his professional and personal email addresses.
Vimal Chandra Pandey (no relation to Bhanu), a researcher at CSIR’s National Botanical Research Institute, edited the book. He told us he would “check in this regard” but did not answer questions asking how the fake references may have been overlooked. Wiley’s page on potential conflicts for editors says they “must refrain from handling manuscripts if they have a COI, like a professional or personal relationship with the authors.”
Other publishers have dealt with similar cases: After retracting a book on machine learning containing non-existent references following our coverage of the case, Springer Nature said in June they would begin issuing expressions of concern for books. In October, research integrity specialist Michał Wójcik wrote a Python script to identify several books with fake citations, which can be a sign of misuse of large language models.
Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on X or Bluesky, like us on Facebook, follow us on LinkedIn, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].
Wiley is taking advantage of AI without doing its due diligence. This is why new processes like AI get such a bad reputation— people don’t exercise appropriate caution when using a new tool.
To take a degenerate example, would you, as a user of a rifle, pick up a machine gun and fire it without precautions due to the fact that you’ve never fired one before and you could kill a lot of people other than the one you’re shooting at? Or even shoot yourself in the foot?
No. The same applies to AI, if the heads of the leading companies are to be believed. Existential risk is not something you just pick up and pull the trigger on.
I guess that Wiley wants the profits from AI without taking the precautions. They should have taken a (cheap) manuscript from India of all places and gone over it with a fine-toothed comb. But nooo.
This would be more convincing if it weren’t so obviously written by GenAI.
As for the silly example the GenAI agent made for you, if someone owned a machine gun and gave it to someone else, or even simply left it unsecured and someone else picked it up, and that other person shot and killed people, that owner would be liable, not just the other person.
Let me know when we start seeing significant judgments being paid out by OpenAI and Anthropic. I won’t hold my breath.
I was recently asked by a publisher to comment on whether a book in my field was worth publishing so investigated the editors alongside the topic. The editor had been senior author on a recent paper which alongside approximately one hundred others had been retracted due to the editor of the journal (a different individual) insisting that all revised manuscripts include him as a co-author as a prerequisite for publication. I forwarded my concerns to the publisher but it did raise concerns about standards in book publishing.
Wiley… It took a year to get this bookchapter retracted. Nonsensical text, irrelevant references, everything plagiarized.
I learned that there are academic editors, the ones mentioned in the book. And there is an editor from the publisher involved. With concerns like this the latter editor asks the academic editors whether something is wrong. And (obviously?) gets to hear everything is fine, lots of novelty in an interesting chapter covering a wide range of topics. This then halts all questions on content. The publisher also runs an automated plagiarism check that finds little. Conclusion: a citation has to be added and all is fine. After many more back-and-forths and independent investigations and what not more the plagiography in the above link did the job.
The same group of editors also produced a book with Springer. Springer is worse than Wiley, mostly not replying. There the “issue” got solved by an online correction, adding some 12 references that were missed when “torturing” the plagiarized texts into something novel. Think of “thoughtful environmental polution”, “downturn of the climate”, and “unspectacular advancement”. Nothing to see here, move on!
https://link.springer.com/chapter/10.1007/978-3-030-94319-6_34