In March, Omar, a biomedical researcher, posted a request in a private Facebook group for researchers. He was seeking to have his name added to a paper in the field of artificial intelligence (AI) and medicine, in exchange for a payment.
Omar is an assistant professor at a Saudi Arabian university. He published 20 papers in 2024 alone, a big jump from 2 the previous year. But it wasn’t enough, says Omar, whose name has been changed to protect his identity. To be promoted, he needed to publish at least ten more papers, in journals in the top half of those ranked by the Web of Science, a citation index. And to meet this demanding requirement, he’d turned to paper mills: fraudulent businesses that deceive publishers so as to arrange authorship, submission and other services related to publications.
Omar wasn’t choosy. The paper could be on any branch of medicine with an AI theme, and he didn’t need to see it before submission. Nor did he mind what position he had in the author list, but he did not want to deal with reviews.
Science’s fake-paper problem: high-profile effort will tackle paper mills
Omar’s circumstances are not unusual: many countries operate a research culture that expects researchers to publish frequently to secure their career in science. Some, like Omar, are unable or unwilling to collect data, conduct analysis and write a paper that might be accepted in a journal. So they turn to organizations online that offer authorship status — for a price. The paper mills that make this their business have various techniques to take advantage of desperate or lazy researchers and to trick publishers: some operate as a marketplace in which extra authorship slots on already–accepted papers are up for grabs. Others take published papers and use AI to tweak text and graphics to escape plagiarism detectors. Still more generate entirely new data sets and text, without the inconveniences of any real laboratory work.
Over WhatsApp, Omar told a reporter for Nature’s careers team, “I am dealing with many writers” as a co-author. Omar wrote that he had already paid for many papers, namechecking co-authors based in India and -Pakistan. He could afford to pay no more than US$400 per paper, he said. Communications he shared indicated that he had paid as little as 76 Saudi riyals (US$20) for one paper. He had marked one Western Union transfer, of 810 Saudi riyals, as an “Education/School fee” on his statement. The reporter contacted Omar after seeing his request on Facebook, and, to explore in more depth why he was offering payment to be included on a paper, posed as the representative of a paper mill.
Omar isn’t the only researcher finding paper mills on Facebook. A separate post offered co-authorship of a paper on medicinal plants and trace elements in return for payment. According to the post, the paper had already been accepted for publication by a respected journal. The reviewers had asked for only minor revisions. Now, the author was seeking other researchers to attach their names to the paper — even though they had not been involved in the research, and would not be needed for substantial revisions.
A US-based plant biologist who was asked by Nature why he had expressed interest in this co-authorship said: “Early-career researchers such as ourselves are under significant pressure to publish, which can sometimes be a challenging aspect of our work.” His Facebook account was then almost immediately deleted.
Ivan Oransky, the co-founder of research–integrity blog Retraction Watch, has repeatedly warned of the industrial-level scale of research misconduct, stressing that paper mills are just a symptom of an academic reward system focused on metrics such as university rankings and quantity of publications. In the absence of any other metric, retraction rates are often used to indicate rates of research misconduct, because retractions often occur only when a publisher decides that a paper is too flawed to be fixed with a correction or an erratum.
Anna Abalkina says that pressure on researchers to publish is causing paper mills to multiply.Credit: Stefanie Loos for Nature
And these pressures are only intensifying, according to Anna Abalkina, who researches academic fraud, including the use of paper mills, at the Free University of Berlin. She points to two reasons: the increasing number of researchers, and the introduction in many countries of requirements to publish in international journals. Paper mills have stepped in to meet this challenge, she adds.
The pressures are particularly intense in some countries. Gengyan Tang, a researcher in education at the University of Calgary, Canada, spent almost a year doing research at a teaching hospital affiliated with a medical university in southwest China. “I found that there was a significant demand for paper mills among nurses,” he reports. “Some paper mills even actively promoted their services within hospital departments by handing out business cards and making direct contact with staff.”
“Early-career doctors and nurses in China are now facing even greater pressure to publish,” says Tang. As evaluation standards grow increasingly stringent, they are expected to produce more papers and publish in higher-impact journals than their predecessors, all while managing heavy clinical workloads, he says. “This struggle to balance research, clinical duties and personal life is one of the key factors driving some of them towards the services provided by paper mills.”
Tang’s concerns around pressures to publish faced by clinical staff are borne out in the data. A Nature analysis of paper retractions from 2014 to 2024, published in February, found that seven of the ten institutions with the highest proportion of retracted research articles were hospitals in China (see Nature 638, 596–599; 2025).
Paper mills explained
Abalkina defines paper mills as “commercial companies that organize on-demand writing of fraudulent academic manuscripts and offer co-authorship of these papers for sale”. For example, International Publisher LLC, a paper mill based in Russia, generated $6.5 million from co-authorship slots offered between 2019 and 2021, according to an analysis by Abalkina1. She estimates conservatively that the paper-mill industry as a whole is worth hundreds of millions of dollars a year (see ‘A case study of paper-mill purchases in Kazakhstan’).
Because papers are often mass–produced, typical signs include familiar text, duplicated images, suspicious e-mail addresses, implausible collaborations1 and irrelevant citations. Another red flag is “tortured phrases” resulting from attempts to hide plagiarism2 (examples include “profound neural organization” instead of “deep neural network”, and “subterranean insect settlement” instead of “ant colony”). Paper mills have also submitted, and occasionally been successful in publishing, plagiarized or duplicated papers.
One tactic is for companies to impersonate well-established researchers when offering papers for sale. In August 2024, a post appeared in a Facebook group selling co-authorship slots, for $200 to $700, on a paper about tracking COVID-19. Supposedly, this paper had already been accepted by the Lancet medical journal. The name on the Facebook account was Paul Robin Krugman, and photos of the Nobel prize-winning economist also appeared in profiles on Facebook and on the social-media app Telegram with the same name. The real Krugman confirmed to Nature’s careers team that this was not him. But not all attempts to use the names of genuine academics are this blatant or easy to detect.
More than 10,000 research papers were retracted in 2023 — a new record
Paper mills offer a bevy of services. A researcher in a bind can find one through word of mouth or by searching online. They might need a paper written from scratch in a particular field, for a particular type of journal. They might already have a draft, but feel that they will have a better chance of acceptance if a third party handles the translation, submission, revision or communication with the journal. They might already have published a paper and want to boost its citations; paper mills often create citation networks to inflate the impact of their papers.
And journal editors willing to accept bribes can accept a paper with little or no scrutiny. Similarly, peer reviewers can give a positive review in exchange for payment. One of Omar’s reviews is just 12 words long.
An unpublished analysis estimates that roughly 2% of all papers submitted to journals in 2022 bore the hallmarks of paper mills, and that the proportion was higher in biology and medicine (see Nature 623, 466–467; 2023). Mass-produced papers in these fields are especially worrying for academic–integrity researchers, because of the damaging effect that this junk science might have on patients.
Easy pickings
In 2021 and 2022, sociologist Xinqu Zhang and criminologist Wang Peng, both at the University of Hong Kong, conducted anonymous interviews with natural-science academics at three elite Chinese universities, as well as obtaining internal documents3.
One faculty member told a research-ethics subcommittee that they had “indeed engaged in academic misconduct”. But they added that if university and faculty leaders “had not implemented such inhumane policies and forced us so harshly” to publish as many articles as possible, they “wouldn’t have resorted to unethical means”.
My ‘woke DEI’ grant has been flagged for scrutiny. Where do I go from here?
This researcher was found out by his colleagues in the faculty, but punishment was lenient: deductions of salary and research funding for that year, without the case being formally reported to others at the university. A committee member handling the case admitted that faculty members admired their colleague’s research, and wanted to avoid negative attention from the rest of the university and the public. A different researcher feared losing his job while dealing with health problems. He told Zhang and Peng that he “had no choice” but to commit research misconduct. He bought access to an official archive and altered the data to support his hypotheses. He then found someone who could help him to write up articles and have them published. “Money is the answer to everything,” he said. “I met all the required criteria in just a few months, and eventually secured a permanent position in the faculty.” These kinds of admission are rare. Authors whose papers have been retracted on suspicion of paper-mill activity tend to remain silent or to deny any participation in research misconduct.
For example, in 2019, Alexandr Litoy, a Moscow–based journalist, purchased an authorship slot for a paper sold by International Publisher. When Litoy, who made the purchase as part of an investigation into paper mills, confronted his two co-authors, who specialize in engineering and chemistry, about why they had contributed to a paper on entrepreneurship, both said that it was their first time using the services of the company. But neither admitted any wrongdoing.
Paper mills continue to make it easy for authors to take shortcuts. Science Publisher, a company registered in Riga, offers co-authorship slots on already written (and sometimes already accepted) papers, in a variety of fields. Prices start from around $550, depending on the field, journal and position in the list of authors. Services are advertised on the company’s website, LinkedIn page and Facebook page.
The company Science Publisher is registered to a residential address in Riga.Credit: Jack Leeming
The company reassures potential buyers that the advertised title and abstract will be rewritten, to ensure that no one finds out that the paper has been purchased. On its website, photos of at least two purported staff members were copied and pasted from a stock-photo site and an entirely unrelated magazine.
A reporter for Nature’s careers team corresponded with Science Publisher about an advertised position as a research-paper author. An employee asked whether she had any ready-made manuscripts, and whether she could provide extra services such as responding to reviewers’ comments.
In the Latvian business registry, Igor Vasilev — whose passport was issued in Volzhsky, in the Russian region of Volgograd — is named as the company’s owner. The Russian business registry indicates that a person of the same name and from the same region was previously chief executive of the company ANO Publishing House “Scientific Review”, which offers services similar to those provided by Science Publisher. Abalkina thinks that the Latvian company is a front for the Russian one.
Science Publisher denies any connection to Vasilev, or even that the company is based in Latvia. A staff member responded to an inquiry from Nature’s careers team with hostility and suspicion, accusing the editor of “commercial espionage” and calling him a “fraudulent impostor”.
In one of its final e-mail communications with the team, a staff member for Science Publisher who described themselves as John Palomarez, director of Academic & Publisher Relations, wrote: “Show your press credentials of Nature on camera? Will you send the proof video of you or any personally [sic] how seeking our Company on behalf of Nature? We’re waiting.”
Afterwards, the address listed on the website was changed from Riga to London, although it retained the Latvian phone number. Reporters representing Nature visited both the Riga and the London address. There were no traces of the company at either location.