Contents
Academy · Research Publication · Community Response to Fraud
Tools, Cultural Change and the Role of the Scientific Community in Combating Fake Papers
Detection software alone has not resolved the problem of fabricated papers. The Declaration on Research Assessment, which is supported by individual and organisational signatories internationally, therefore targets the evaluation incentives that make fabrication worthwhile, alongside shared screening infrastructure that detects fraud which individual technical tools would miss. This article examines both lines of response and the evidence for each.
Responding to fabricated and manipulated research requires two distinct, complementary categories of response: technical detection tools that identify fraud already in the publication pipeline, and cultural or structural reforms that reduce the underlying incentive to commit it. Treating either category as sufficient on its own has repeatedly proven inadequate: software cannot correct an evaluation system that rewards precisely the behaviour that fraud is designed to simulate, and cultural appeals alone cannot detect sophisticated, industrial-scale fabrication engineered to pass a cursory human review.
Learning objectives
- The detection technologies that publishers have developed to identify fabricated and manipulated submissions
- Why editors-in-chief and cross-publisher coordination have become a distinct line of response
- What the research assessment reform movement targets, and the current extent of its adoption
Common misconceptions
- “Improved detection software is, by itself, the solution to the problem of fake papers.” Research on the root causes of scientific fraud consistently identifies evaluation systems weighted by publication count as a structural cause. For this reason the Declaration on Research Assessment, which is supported by individual and organisational signatories internationally, exists as a parallel reform track alongside detection technology, rather than detection tools alone being presented as a sufficient response.
Detection technologies developed by publishers
Several major publishers and industry groups have developed dedicated detection software aimed at the two most common markers of fabrication: manipulated or duplicated images, and generated or heavily templated text. Guillaume Cabanac and Cyril Labbé’s research on algorithmically generated papers and on “tortured phrases”, the garbled synonym substitutions that automated paraphrasing tools tend to introduce, provided the field with a concrete, searchable signature for which screening software can scan across large submission volumes1. The STM Integrity Hub, developed by the International Association of Scientific, Technical and Medical Publishers (STM) to pool detection capability across member publishers, screens submissions against a shared database of known paper-mill indicators and image-manipulation signatures, addressing the fact that fraud distributed thinly across many journals is invisible to the tools of any single publisher working in isolation2. Jana Christopher’s practical guidance for editors and reviewers on detecting image and data manipulation complements this software layer with a layer of trained human review, since some markers of manipulation still require expert visual judgement that automated tools do not yet reliably replicate3.
The importance of cross-publisher coordination
Because paper mills exploit the fact that no single journal receives enough submissions from one source to detect a pattern, coordination among editors-in-chief across journals, and even across competing publishers, has become a distinct and necessary response alongside individual detection tools. The joint Committee on Publication Ethics (COPE) and STM research project on paper mills recommended that publishers share information about suspect submissions and authors across companies, a step that requires active institutional coordination rather than improved software at the level of individual journals4. The same report stated explicitly that shared blacklisting raises genuine data-protection and fairness concerns, since falsely flagging a legitimate researcher on the basis of a shared, cross-publisher record carries a different and larger reputational cost than an internal error by a single journal; the coordination itself therefore requires careful governance design, not merely technical commitment4.
Targets of the research assessment reform movement
The Declaration on Research Assessment (DORA), launched in 2012, targets the evaluation metrics that research on the root causes of fraud has repeatedly identified as part of the problem: the use of publication count and venue prestige, rather than research quality and broader contribution, as the basis for hiring, promotion and funding decisions5. DORA is supported by individual and organisational signatories internationally, including universities, funders and professional societies, and represents one of the largest coordinated efforts within the research community to change the underlying incentive structure rather than only addressing its symptoms after the fact5. The 2017 consensus report of the National Academies of Sciences, Engineering, and Medicine on fostering research integrity similarly recommended that institutions adopt clearer, less metrics-dependent evaluation criteria as part of a broader strategy that also includes research-integrity training and transparent misconduct-investigation processes6.
Practical contributions of individual researchers and reviewers
Institutional and technological reforms operate at a scale that individual researchers cannot change, but the literature on community responses also identifies specific, realistic actions available to an individual reviewer, editor or co-author. Christopher’s guidance for reviewers recommends specific, learnable checks: comparing the methods section of a manuscript with the author’s previously published work for suspicious structural identity, checking whether reported sample sizes and statistical results are internally consistent, and flagging image panels that appear reused or inconsistent with the rest of the figure in resolution or background noise. These are concrete tasks that any trained reviewer can perform without specialised forensic software3. Teixeira da Silva and Dobránszki’s analysis of post-publication peer review also highlights the role of readers after publication: public comment platforms and post-publication review venues have brought to light a meaningful share of documented fraud cases because an engaged reader noticed an inconsistency that the formal pre-publication review process had missed; the collective vigilance of the research community after publication therefore functions as a real, if informal, extension of the formal detection system7. Co-authors also carry a specific documented responsibility: COPE guidance holds every listed author jointly accountable for the integrity of a paper, which supports co-authors reviewing the underlying data themselves before publication, rather than relying entirely on the assurances of the lead author, as one of the simplest and most immediate safeguards available before a paper is submitted4. Byrne and Christopher’s companion analysis of gene-function paper mills, and Candal-Pedreira and colleagues’ tracing of retracted papers to specific mills, both note that several of the cases studied were first flagged by attentive individual readers or reviewers rather than by any institutional tool, indicating that trained human attention remains a meaningful layer even as detection technology improves89.
Limitations of the current evidence base
Jaime Teixeira da Silva and Judit Dobránszki’s analysis of post-publication peer review and correction practices noted that much of the current understanding of how well these combined interventions, involving both tools and culture, actually perform still rests on individual case studies, anecdote and relatively small samples, rather than large, systematic, controlled evaluations of intervention effectiveness7. This is a genuine limitation for any attempt to decide with confidence how much weight to place on a single intervention (detection software, cross-publisher coordination or DORA-style evaluation reform), and it supports treating current best practice as a reasonable, evidence-informed combination of plausible measures rather than a fully proven, optimised system. The evidence gap is itself useful information for those designing an institutional response: it supports adopting a reasonable combination of the measures described in this article (detection software, cross-publisher coordination, evaluation reform and individual researcher vigilance) while monitoring and reporting on how well each measure performs at a given institution or journal, rather than assuming that any single widely recommended intervention has already been proven optimal elsewhere. Such internal monitoring is itself a contribution to the broader evidence base that the field still requires, and journals or institutions willing to publish or share their own before-and-after data on the effect of a specific intervention on detection rates or retraction counts would meaningfully narrow the gap that Teixeira da Silva and Dobránszki identify. In the meantime, the absence of a single, definitively proven intervention is not a reason for inaction; it is a reason to implement several plausible, evidence-informed measures concurrently, since the two broad categories of response, detection technology and incentive reform, target different parts of the problem and cannot substitute for one another, regardless of the eventual findings of more rigorous comparative evidence. A journal or institution starting from a limited base may reasonably prioritise whichever of the two layers it is best positioned to act on immediately, for example by joining existing shared detection infrastructure rather than building bespoke tools, or by adopting an existing framework such as DORA rather than drafting new evaluation criteria, while treating the other layer as a deliberate medium-term goal rather than deferring both indefinitely in the absence of a fully proven solution. The case studies and reform efforts surveyed in this article suggest that neither layer alone has been sufficient in the documented instances: journals that invested heavily in detection technology without addressing evaluation incentives have continued to encounter new paper-mill activity, while institutions that reformed evaluation criteria without improving detection have still missed sophisticated fabrication engineered to pass cursory review.
Key points
Key points
Addressing fabricated research requires both layers to operate together: detection technology and cross-publisher coordination identify fraud that has already been submitted, while reform of evaluation systems, such as DORA, which is supported by individual and organisational signatories internationally, addresses the underlying incentive that makes fabrication worthwhile. The evidence on the performance of any single intervention remains thinner than the scale of the problem warrants, so current best practice should be regarded as a reasonable combination rather than a fully validated solution.
Practical application
Practical application
- Both detection layers, automated screening and trained human review, should be used. Automated tools detect patterns such as tortured phrases and image duplication at scale, but some markers of manipulation still require expert visual or editorial judgement that software cannot yet reliably replicate.
- Participation in cross-publisher detection infrastructure is advisable where the relevant role permits it. Individual journals cannot observe the cross-journal patterns that characterise paper-mill fraud, so shared systems such as the STM Integrity Hub are necessary for detecting distributed fraud.
- Evaluation criteria that do not reward raw publication count should be promoted. Institutional or journal assessment criteria should be reviewed against the principles of DORA, since evaluation weighted by publication count is a documented structural driver of the pressures that lead to fraud.
Frequently asked questions
No. Research on the root causes of fraud consistently identifies evaluation incentives, not only gaps in detection, as a structural driver, which is why reform efforts such as DORA exist as a parallel track alongside detection technology rather than technology being treated as a sufficient standalone solution.
It is shared detection infrastructure, developed by the International Association of Scientific, Technical and Medical Publishers, that screens submissions against a pooled database of known fraud indicators; it requires cross-publisher cooperation because paper mills distribute their output thinly across many journals specifically to avoid detection by the tools of any single publisher.
DORA is supported by individual and organisational signatories internationally, including universities, funders, and professional societies, and is one of the largest coordinated efforts to shift research evaluation away from publication-count and venue-prestige metrics.
A joint COPE and STM research project on paper mills specifically identified data-protection and fairness concerns, since a shared blacklist could falsely flag a legitimate researcher, with larger reputational consequences than an error confined to one journal; such sharing therefore requires careful governance design rather than unrestricted exchange.
Not yet at the scale that the problem warrants. Analyses of post-publication review and correction practices note that much current understanding still rests on individual case studies and relatively small samples rather than large, controlled evaluations of intervention effectiveness.
References
- Cabanac G, Labbé C, Magazinov A. Tortured phrases: A dubious writing style emerging in science. arXiv:2107.06751. read 10 October 2026. View on arXiv
- STM (International Association of Scientific, Technical and Medical Publishers). STM Integrity Hub. read 10 October 2026. View the STM Integrity Hub
- Christopher J. The raw truth about paper mills. FEBS Letters. 2021;595(13):1751-1757. read 10 October 2026. View on FEBS Letters
- Committee on Publication Ethics and STM. Paper mills research project report. read 10 October 2026. View the COPE/STM paper mills report
- Declaration on Research Assessment (DORA). read 10 October 2026. View DORA
- National Academies of Sciences, Engineering, and Medicine. Fostering Integrity in Research. The National Academies Press, 2017. read 10 October 2026. View the National Academies report
- Teixeira da Silva JA, Dobránszki J. Problems with traditional science publishing and finding a wider niche for post-publication peer review. Accountability in Research. 2015;22(1):22-40. read 10 October 2026. View on Accountability in Research
- Byrne JA, Christopher J. Digital magic, or the dark arts of the 21st century-how can journals and peer reviewers detect manuscripts and publications from paper mills? FEBS Letters. 2020;594(4):583-589. read 10 October 2026. View on FEBS Letters
- Candal-Pedreira C, Ross JS, Ruano-Ravina A, Egilman DS, Fernández E, Pérez-Ríos M. Retracted papers originating from paper mills: cross sectional study. BMJ. 2022;379:e071517. read 10 October 2026. View on BMJ
- Else H, Van Noorden R. The fight against fake-paper factories that churn out sham science. Nature. 2021;591:516-519. read 10 October 2026. View on Nature
Every source above is a peer-reviewed article, a publisher’s record of one, or a standard issued by the body that owns it. Each carries the date it was last read. Where a standard moves, this article is revised rather than left up with a note.
How to cite this article
Pick your referencing style and copy it. Your university or target journal may vary the punctuation — check its guide if it has one.
Pubrica Academy. (2026, October 10). Tools, Cultural Change and the Role of the Scientific Community in Combating Fake Papers. Pubrica. https://pubrica.com/academy/study-guide/publication/essential-tools-for-fighting-fake-papers/
Pubrica Academy. “Tools, Cultural Change and the Role of the Scientific Community in Combating Fake Papers.” Pubrica, 10 Oct. 2026, pubrica.com/academy/study-guide/publication/essential-tools-for-fighting-fake-papers/.
Pubrica Academy (2026) Tools, Cultural Change and the Role of the Scientific Community in Combating Fake Papers. Pubrica. Available at: https://pubrica.com/academy/study-guide/publication/essential-tools-for-fighting-fake-papers/ (Accessed: 15 October 2026).
Pubrica Academy. “Tools, Cultural Change and the Role of the Scientific Community in Combating Fake Papers.” Pubrica. October 10, 2026. https://pubrica.com/academy/study-guide/publication/essential-tools-for-fighting-fake-papers/
Pubrica Academy, ‘Tools, Cultural Change and the Role of the Scientific Community in Combating Fake Papers’ (Pubrica, 10 October 2026) <https://pubrica.com/academy/study-guide/publication/essential-tools-for-fighting-fake-papers/> accessed 15 October 2026.
Pubrica Academy. Tools, Cultural Change and the Role of the Scientific Community in Combating Fake Papers. [Internet]. Pubrica; 2026 Oct 10 [cited 2026 Oct 15]. Available from: https://pubrica.com/academy/study-guide/publication/essential-tools-for-fighting-fake-papers/
Pubrica Academy, “Tools, Cultural Change and the Role of the Scientific Community in Combating Fake Papers,” Pubrica, Oct. 10, 2026. [Online]. Available: https://pubrica.com/academy/study-guide/publication/essential-tools-for-fighting-fake-papers/
@misc{/la_tools_cultural_change_fighting_fake_papers_2026,
author = {{Pubrica Academy}},
title = {Tools, Cultural Change and the Role of the Scientific Community in Combating Fake Papers},
howpublished = {Pubrica},
year = {2026},
month = {oct},
url = {https://pubrica.com/academy/study-guide/publication/essential-tools-for-fighting-fake-papers/},
note = {Accessed: <span data-cite-acc="iso">2026-10-15</span>}
}This article is written to be learned from. Cite it if you are referring to what it says; if you are using the evidence behind it, cite the sources listed above instead.
Two-Layer Anti-Fraud Response Checklist
This checklist assesses whether the anti-fraud approach of a journal or institution covers both layers.
[ ] Automated detection tools in place for image duplication and generated/templated text [ ] Trained human reviewer layer for manipulation markers software does not reliably catch [ ] Participation in or access to cross-publisher detection infrastructure (e.g., STM Integrity Hub) [ ] Evaluation criteria reviewed against DORA principles to reduce publication-count-weighted incentives [ ] Research-integrity training provided to editors, reviewers, and early-career researchers [ ] Transparent, COPE-compliant process for investigating and acting on suspected misconduct
Pubrica AcademyWritten and maintained by Pubrica’s subject matter experts. Articles are revised when the underlying standards change, and each article states the date on which it was last reviewed.
How Pubrica can help
Pubrica’s research integrity consultancy assists journals and institutions in establishing both the detection and the incentive-reform layers required to respond to fabricated research at scale.