Today publishing no longer means simply writing a good article. It means choosing the right journal, navigating between open access and transformative agreements, understanding metrics, managing the peer review process, safeguarding scientific integrity, and dealing with increasingly complex editorial policies. Authors, editors, and researchers find themselves at the center of a rapidly evolving system, where rules change quickly and strategic decisions can make a real difference.

This monthly newsletter was created to provide practical tools and clear analysis. Each issue is designed to offer context, guidance, and actionable insights.

Not just news, but ways to interpret it. If you work in research or academic publishing, this space is for you.

Welcome aboard!

The Journals That Prey on Publish-or-Perish

Somewhere in an inbox near you sits an email that begins, "Greetings Eminent Researcher," and promises publication within seven days, for a fee. It's an invitation to a predatory journal — and academia has spent the better part of two decades trying to figure out what to do about it.

What They Are

Predatory journals mimic the trappings of legitimate scholarly publishing — an editorial board, a peer review process, an impact metric — while gutting the substance behind each one. Peer review, if it happens at all, is often cursory or entirely fabricated. Editorial boards sometimes list academics who never agreed to serve, or who don't exist. The "impact factor" displayed on the homepage is frequently a made-up number, or one purchased from a disreputable indexing service with no relationship to the real Journal Citation Reports.

The business model is simple: charge authors an article-processing fee, publish almost anything submitted, and repeat at scale. Some operations run hundreds of journals simultaneously, harvesting fees from thousands of authors a year.

Why Researchers Fall In

It's tempting to imagine predatory journals as preying only on the naive. The reality is more complicated. Early-career researchers under pressure to publish quickly face genuine incentives to take a shortcut. Scholars in countries where funding or promotion is tied to raw publication counts, rather than journal prestige, may rationally calculate that a fast, cheap publication satisfies institutional requirements even if it does little for their actual reputation. And predatory publishers have gotten better at camouflage — cloning the look of legitimate journal websites, adopting near-identical titles to established ones, and buying advertising space that makes them appear in the same search results as reputable outlets.

The Cost

The damage isn't confined to individual CVs. Predatory journals pollute the scientific record. Unverified findings — some outright fraudulent, others simply never checked — enter citation databases and occasionally make their way into systematic reviews, treatment guidelines, or policy documents. In fields like medicine, that's not a hypothetical harm.

There's also an institutional cost. Universities and funders that unknowingly count predatory publications toward tenure decisions or grant outputs end up rewarding volume over rigor, quietly degrading the incentive structure for everyone.

Fighting Back

The response has been a mix of blacklists, whitelists, and cultural change. Beall's List, a widely cited (if controversial) inventory of suspected predatory publishers, though its original maintainer stopped updating it years ago, spawned successors and a broader movement toward journal-vetting tools like DOAJ (Directory of Open Access Journals) and Cabells' Predatory Reports.

But technology alone won't solve a problem rooted in incentives. As long as institutions measure scholarly output by counting papers rather than reading them, there will be a market for journals willing to publish anything for a price. The most durable fix isn't a better blacklist — it's a publishing culture that rewards quality over quantity, and gives researchers enough time and support to avoid needing a shortcut in the first place.

"Review mills" put peer review under strain

A Nature investigation published on 3 August 2026 shines a light on "review mills": researchers who recycle boilerplate referee reports and pad them with irrelevant citation demands, apparently to boost their own or their collaborators' citation counts. The term comes from marketing researcher Maria Ángeles Oviedo-García, whose 2024 Scientometrics paper flagged 263 reviewer reports showing signs of this behaviour.

The scale is still modest but real. At Institute of Physics Publishing (IOPP), 0.05% of reviews received in 2025 turned out to be fully identical, and another 0.3% overlapped by 80% or more. A separate look at nearly 150,000 reports across MDPI, PeerJ, The BMJ and computer-science journals found 0.5–1.3% highly similar to others.

It appears that most people involved in review mills seek to boost their citations numbers. The citation math can get striking. One engineer's most-cited paper has racked up 748 citations, 79 of them self-citations, and around 16% of his overall citations are self-generated. A retracted paper he co-authored cited his own work 123 times. Investigators have also traced organised "cartels," including a group of nine Italian oncologists who used shared review templates to secure citations across 170 articles.

What makes this hard to police is that reviewers often ask authors to cite a collaborator's paper rather than their own, and most peer-review reports are never made public in the first place. Publishers including MDPI, PLOS ONE and IOPP are now rolling out duplicate-review checkers and reviewer bans, but authors caught in between say the reputational fallout still lands on them rather than on the reviewers or the journals.

SOURCE

AI use in manuscripts is rising, but authors still barely disclose it

Two companion studies published in JAMA over the past few months paint a picture of AI's quiet creep into scientific writing — and how little of it authors actually admit to.

The first, covering 105,538 manuscripts submitted to 13 JAMA Network journals over 27 months (August 2023 to October 2025), found that 3.3% overall declared AI use — and that the disclosure rate climbed steadily, from 1.71% at the start of the study period to 5.97% by the end. Most disclosed use was for language correction and refinement (about two-thirds of cases), followed by statistical modelling, other data analysis, and manuscript drafting. Disclosure was notably more common in Viewpoints and Letters to the Editor than in original research articles, more common among authors from non-English-speaking countries, and much more common in manuscripts that were later withdrawn or rejected than in those ultimately accepted.

The second study, led by Isamme AlFayyad and colleagues, looked at 25,114 manuscripts submitted to 49 BMJ Group journals between April and November 2024, after BMJ made AI disclosure mandatory. Only 5.7% of authors disclosed any AI use — far below the 28–76% that earlier self-report surveys had suggested was actually happening. Disclosure did creep up over the study period, from 4.5% in April to 7.3% by October, hinting that norms may be shifting, but the gap between reported behaviour and survey-estimated reality remains wide.

Read together, the two studies suggest that mandatory disclosure policies are still catching only a fraction of real AI use in manuscript preparation — a transparency problem publishers are only beginning to grapple with.

SOURCE

SOURCE