In recent years, paper mills have become one of the most discussed threats to the integrity of the scientific record.
Publishers have retracted thousands of papers linked to organised manuscript production. Investigations have revealed networks that generate papers at industrial scale. Entire businesses now exist whose sole purpose is to produce manuscripts and sell authorship positions to researchers who need publications.
Most discussions focus on detection. Journals develop tools to identify manipulated images, fabricated peer review, or unusual citation patterns. Retractions follow and the publishers explain that safeguards are being strengthened.
These efforts are important. Fraudulent research should not remain in the literature and, if detected, the articles should be retracted.
But focusing only on detection raises a more fundamental question. Why does a market for manufactured research exist? Understanding that question requires looking beyond individual misconduct and examining the rewards for publishing. If we can answer that question, we address the problem at source, rather than trying to resolve an ever-recurring problem.
The industrialisation of academic papers
Paper mills did not suddenly appear.
They emerged gradually as academic publishing became more closely tied to career advancement. In many universities, publication is the most visible signal of academic productivity. Promotion decisions, grant applications, and institutional rankings often rely heavily on publication metrics.
When professional success depends on producing papers, demand for publications increases. Under this publish or perish pressure, some researchers begin looking for ways to obtain publications quickly.
Paper mills feed that demand.
The services offered vary. Some organisations provide ghostwritten manuscripts. Others sell authorship positions on papers that have already been written. In some cases, the paper itself is constructed from templates that combine stock figures, fabricated data and automated text generation.
What matters is not the exact method, but the economic model. Paper mills produce something that universities appear to value: publications. In scientific terms these papers may be worthless, but within evaluation systems they can still function as indicators of productivity.
In effect, paper mills manufacture the indicators that institutions use to evaluate researchers. Many universities will not have the time, expertise or resources to validate these papers, so they just get assimilated in the pool of legitimate papers. Moreover, if these paper mill publications are presented in the right way, especially if they can be published in what appear to be legitimate journals, then universities may sometimes prefer not to examine the issue too closely.
Where the demand comes from
It is easy to assume that researchers who use paper mills are simply attempting to cheat the system. In some cases that may be true. But the broader picture is more complicated.
Many academic systems contain strong incentives to publish. Promotion rules may require a fixed number of journal articles. Financial bonuses may be tied to papers indexed in certain bibliographic databases. National research evaluation exercises often reward publication volume or journal rankings. Under these conditions, publishing becomes more than scholarly communication. It becomes a requirement for career survival and advancement.
Most researchers respond to these pressures in legitimate ways. They conduct research, collaborate with colleagues and submit papers through normal editorial processes.
But when the pressure increases and the reward structure is rigid, alternative markets begin to appear. Paper mills exist because there is demand for what they provide.
Why detection alone cannot solve the problem
Publishers have invested considerable effort in identifying paper mill products.
Image manipulation detection has improved. Peer review processes are scrutinised more closely. Statistical anomalies can sometimes reveal fabricated datasets. Journals now retract papers linked to suspicious networks. These efforts have exposed many fraudulent publications.
Yet the problem continues and, if anything, is getting worse.
Part of the difficulty is that paper mills adapt quickly. When one technique becomes detectable, another appears. As detection methods improve, the production methods evolve. We are effectively in an arms race. Detection methods become more sophisticated, but paper mills also become more innovative in avoiding them.
There is also a deeper issue. Detection addresses the supply of fraudulent papers. It does not address the demand that drives the market. As long as researchers are rewarded primarily for producing publications, the incentive to publish papers by any available means will remain.
In other words, paper mills are not simply exploiting weaknesses in publishing workflows. They are responding to the incentives embedded in research evaluation systems.
The role of universities and research institutions
This is where the conversation often becomes uncomfortable.
Publishers and journals are visible actors in the scholarly communication system. When fraudulent papers appear, attention naturally focuses on editorial processes and peer review safeguards.
Universities and research institutions receive less scrutiny.
Yet many of the incentives that create demand for paper mills originate within academic organisations themselves. Promotion rules, performance evaluations and financial rewards can unintentionally encourage publication volume over research quality.
If institutions reward publication counts without examining how those publications were produced, the signals become vulnerable to manipulation. This does not mean that universities support misconduct. But it does mean that institutional structures can unintentionally create environments where paper mills flourish.
Rethinking how research is evaluated
Addressing the paper mill problem requires that we go beyond ever better detection tools. It requires reflection on how academic performance is measured.
Many institutions are already experimenting with new approaches. Narrative CVs emphasise research contributions rather than publication counts. Some funding agencies ask applicants to highlight a small number of significant outputs rather than listing every paper.
These changes attempt to shift attention away from quantity and toward substance.
The transition is not simple. Publication metrics remain attractive because they are easy to measure and compare, and largely automated. Changing evaluation systems requires institutional courage and collective agreement across the research community.
But if the incentives remain unchanged, the underlying pressures will persist.
A symptom of a larger system
Paper mills are often described as a corruption of scholarly publishing. In some ways they are. Yet they can also be understood as a symptom of a broader system in which academic success is closely tied to measurable outputs.
When outputs become the primary signal of achievement, markets tend to appear that supply those outputs. Paper mills are one result of this market pressure. Understanding this does not excuse misconduct. Fraudulent research damages the credibility of science and wastes the time of researchers who rely on the scientific literature.
But recognising the structural forces involved helps explain why the problem has grown so quickly.
If the academic community wants to reduce the influence of paper mills, it may need to examine not only how papers are produced but also how they are rewarded.
A question for the research community
The discussion around paper mills often focuses on publishers and editors. Perhaps it should also focus on universities and the incentives they offer. If publication metrics create the demand for manufactured papers, can the integrity of the literature be protected without rethinking how academic success is measured?
Following the series
This newsletter examines the changing structure of scholarly publishing and the incentives shaping research behaviour.
Future editions will explore topics such as authorship markets, citation manipulation and the growing industrialisation of paper production.
If you would like to follow the series, please subscribe. LinkedIn will notify you when each new edition is published.
About the author
Graham Kendall is an experienced researcher and former university Vice-Chancellor with more than 300 peer reviewed publications and over 20,000 citations. His work spans artificial intelligence, optimisation and research systems, and he has spent many years examining the structures that shape academic careers and scholarly publishing.
Through the Publishing with Integrity initiative, he writes and speaks about the evolving challenges facing academic publishing, including paper mills, authorship markets, citation manipulation and the impact of artificial intelligence on research communication.
He also advises universities, publishers and research organisations on issues related to research integrity, evaluation systems and the governance of academic publishing. If your institution is reviewing its research policies or seeking to strengthen its integrity framework, he welcomes the opportunity for conversation.