A paper mill is an organization that mass-produces and sells scientific manuscripts. These manuscripts are often based on fabricated data or plagiarized content. Their sole purpose is to generate publications for authors who are under pressure to publish. This is no longer a marginal phenomenon. Today, it is one of the most serious threats to the reliability of the scientific literature.
A Rapidly Growing Threat
Recent bibliometric studies point to the same conclusion. The number of articles originating from paper mills has increased sharply since 2020. A joint report by COPE and STM estimated that suspicious submissions account for around 2% of all submissions across journals. In journals specifically targeted by paper mills, this figure can rise to as much as 46%.
Another worrying sign is that the overall retraction rate remains far below the estimated scale of the problem. It stood at around 0.2% in 2025, compared with an estimated actual rate of close to 2%. In other words, a large proportion of fraudulent articles may never be detected.
Warning Signs to Look Out For
Paper mills leave recognizable traces if you know where to look. Here are the main warning signs to watch for.
• Suspicious Writing Patterns
Repetitive wording sometimes appears across multiple manuscripts submitted by different authors. These papers may also contain “tortured phrases”: automatically generated paraphrases designed to bypass plagiarism-detection software. For example, “artificial intelligence” might become “synthetic intelligence.”.
• Unusual Spikes in Submissions
A special issue or niche topic may suddenly receive an unusually large influx of manuscripts. These spikes often occur through guest editors who may be less experienced in editorial screening.
• Inconsistent Author Networks
Co-authors’ affiliations may have no apparent connection to the research topic. Likewise, reviewers suggested by the authors may be fictitious, or their email addresses may actually be controlled by the authors themselves.
• Problematic References
Paper mills often fabricate entire bibliographies. They may include citations to nonexistent sources or to articles that have already been retracted. This is generally one of the easiest warning signs to verify automatically.
Why Manual Checks Are No Longer Enough
Historically, detection relied on the vigilance of editors-in-chief and peer reviewers. But given the scale at which paper mills operate today, this manual approach is reaching its limits. The volume of manuscripts, the growing sophistication of AI-generated text, and the subtlety of citation networks make manual detection slow and incomplete.
This is precisely where the integrity of peer review is at stake. An editorial team without a systematic verification tool therefore takes a disproportionate risk compared with the amount of time available to assess each manuscript.
Moving Toward Tool-Assisted Detection
COPE recommends a collective response to the problem. This includes sharing information between publishers and developing automated detection tools. Systematic verification of bibliographic references is one of the first lines of defense.
An inconsistent bibliography, citations that cannot be found, or an unusually high concentration of cross-citations within a small group of authors can often be the first visible indication that a manuscript was produced by a paper mill. Integrating automated checks as soon as a manuscript is submitted does not replace editorial judgment. Instead, it frees up time to focus on what matters most: the scientific quality of the submitted research.
LibScience develops an automated bibliographic reference verification solution for scientific journals and publishers, including the detection of retracted articles, automatic DOI enrichment and contextual citation relevance checking. Discover LibScience