PMC and a Withdrawn Preprint

Remember when it meant something to have a paper retracted or withdrawn?

Oh, those quaint old days of a few years ago . . .

Here are two stories illustrating how meaningless withdrawing anything has become thanks to our mindless embrace of “more information is better” thinking, libertine platforming tech, and irresponsible registration and indexing practices.

1. Preprints.org recently withdrew a preprint only to have it reposted on Zenodo with just a disclaimer atop it. It’s a preprint designed to foment vaccine hesitation, fear-mongering to parents of young infants. It comes from Children’s Health Defense, where RFK, Jr., cut his vaccine conspiracy teeth, with the notorious Brian Hooker and Karl Jablonowski as the authors. Neither has any qualifications in vaccine science.
2. Another preprint withdrawn by Preprints.org in July 2025 (our coverage of its posting here, and of the September withdrawal here) was republished in a “gaslight” journal in December 2025, and has now been indexed and published in full-text on PubMed Central. Of course, it has to do with mRNA vaccines causing cancer and comes from the McCullough Foundation, another reliable source for unreliable vaccine fear-mongering. They were the ones who caused us to coin the term “gaslight” journals, after all.

The fact that a commercial preprint service has higher standards than a DOI-registered and PMC-indexed journal is depressing, showing how deeply the infrastructure of scientific publishing — including CrossRef and the NLM — has been corrupted by OA and cyberlibertarian cultures. Our registration and indexing systems accept anything. They apparently have few if any standards left, and haven’t built any cross-checking systems into their tech stacks.

Comparing the withdrawn July 2025 preprint with the December 2025 “gaslight” journal version indexed and reproduced in full-text on PMC shows that little changed — the “gaslight” journal version has more metadata, looser statistical boundaries, three fewer authors, and some conflicts of interest statements and journal-esque details. Otherwise, it’s the same garbage, just in a clean can.

• You can still find the withdrawn preprint on ResearchGate. I’m not linking to it, however.

The automated box of confusion is running very efficiently, PMC is a growing hotbed of OA misinformation, and an article withdrawal now means nothing.

This also means that AI and LLM systems are probably ingesting this misinformation even as we speak . . . you know, to make it permanent while obscuring its provenance and terrible backstory.

We’ve really made a mess of things . . . maybe it’s not such a bad thing that the NLM LinkedIn account is being shut down. Could be a good time to stop relying on the NLM overall.