bioRxiv has posted 6,490 single-author preprints. 668 of them this year.
On August 13 they told me, in writing, that not one of those should technically be there. Here is the rule they sent:
we require that a Principal Investigator, affiliated with an established research institution, under whose guidance the work was performed, submit the manuscript and be listed as the corresponding author.
Read that again. It isn’t “have an affiliation.” It’s a lab boss. Somebody senior, at a place they rank, who submits your paper for you and takes the corresponding-author line on work you did alone.
If you’re a sole author you don’t have one of those. Nobody does. That’s what sole author means.
Where the rule isn't
I went looking for it. I spent a large amount of time looking, but turned up empty-handed.
Their submission guide lists three conditions. Relevant field. Unpublished. All authors consent. That’s the list. The word “affiliation” is not on that page. Not once. It’s not in the FAQ either.
So I asked where it’s written. They sent me a blog post from May by their CSO and CEO. The closest thing to a rule in it is that openRxiv “requires authors to supply an affiliation.”
“That’s bullshit”, I thought. I supplied one. The Institute for Complexity Science and Advanced Computing. On the author line and submitted from its own domain. Registered US research organization, Crossref member, its own DOI prefix, ORCID-gated editorial workflow.
“Principal Investigator” appears nowhere on that page. Nowhere. I also asked whether ICSAC had actually been evaluated before being judged insufficient. Of course, they didn’t answer that part.
Their own form never asks
Every field the submission form collects. Category. Subject area. Title. Author name and institution, one free-text box. External data. Competing interests. Funding. License. Abstract.
No PI field. Nothing that checks the institution against anything.
How do I know nothing checks it? “Michiagan State University” is on their server right now. So is “University Colloge Cork.” Sic, both of them.
What they actually post
This year alone, nine preprints declare their affiliation as some version of independent researcher. Four say exactly that. Two more just simply say N/A.
Two days before I submitted, they posted a single-author paper affiliated to a private company. A floating solar developer, with a business registration twelve months old.
That paper is microbiology and it belongs on bioRxiv. I’m not naming it. Its author did nothing wrong and that isn’t the story. The affiliation is. bioRxiv says it needs one so an organization can adjudicate ethical disputes about the research. A solar company is not going to adjudicate a fight about antibiotic resistance data.
Those papers are fine. bioRxiv was right to post them. That’s the whole point. The rule they quoted at me isn’t the standard they run.
Yes, the AI flood is real
It is. It’s real. I’m not going to pretend the slop problem is invented.
But an affiliation check never opens the paper. Generated garbage from a .edu sails right on through. Good work on the “wrong” letterhead or “wrong” affiliation dies totally unread. It filters people, not papers. And again, it’s bullshit.
Why I know any of this
They rejected mine in 18 hours. The rejection doesn’t mention the science. Not the methods, not the data, nothing past the author line.
I submitted from an institutional email with the institution on the author line. Not enough. They won’t say what would be.
Who decides what counts as established
Their own about page lists who funds them. Chan Zuckerberg Initiative. Cold Spring Harbor Laboratory. The Sergey Brin Family Foundation. Caltech. CNRS. Fred Hutchinson. Imperial. MIT. Stanford. Edinburgh. Washington.
Their board secretary is the president of Cold Spring Harbor. Their chair and their chief science officer founded bioRxiv out of Cold Spring Harbor Laboratory Press. Their CEO used to run Dryad, which that same May blog post names as one of their trust partners.
I’m not alleging anyone did anything improper. I’m saying the body that decides which institutions count is funded and governed by the institutions that count.
What happens to the paper
It goes up as a poster at The Science of Consciousness 2026 San Diego in October. A committee read the abstract and took it.
The preprint posts somewhere that writes its rules down first.
Check it yourself
The counts come from OpenAlex, bioRxiv source S4306402567, filtered to authors_count:1. They agree with bioRxiv’s own API to within 0.3 percentage points. The page they cited is archived by the Wayback Machine on 2 June and 9 June, both before I submitted.
The ask
Run the query yourself. One tab. It replicates or it doesn’t. If a number here is wrong, tell me and I’ll happily fix it in public.
And if a preprint server has bounced you over affiliation, I want to hear about it. There’s no public record of how often this happens or to who, which is exactly why it keeps happening quietly. Reply to this email, or write to research@nathanthornhill.com.
~ If you know someone who may enjoy reading this article, please share ~
OpenAlex, 6,490 sole-authored: https://openalex.org/works?page=1&filter=primary_location.source.id%3AS4306402567,authors_count%3A1
OpenAlex, 668 in 2026: https://api.openalex.org/works?filter=primary_location.source.id:S4306402567,authors_count:1,publication_year:2026
bioRxiv own API, cross-check: https://api.biorxiv.org/details/biorxiv/2026-08-01/2026-08-13
cited blog post, live: https://openrxiv.org/trust-signals-and-accountability/
cited page archived 2 June: https://web.archive.org/web/20260602165923/https://openrxiv.org/trust-signals-and-accountability/
cited page archived 9 June: https://web.archive.org/web/20260609210034/https://openrxiv.org/trust-signals-and-accountability/
submission guide archived: https://web.archive.org/web/20260719090001/https://www.biorxiv.org/submit-a-manuscript
FAQ archived: https://web.archive.org/web/20260802203550/https://www.biorxiv.org/about/FAQ
openRxiv about, funders + board: https://openrxiv.org/about/