Is philanthropy the future of academic publishing?

Normal people are often incredulous when you tell them about the model of academic publishing. Conversations tend to go something like this:

“Oh, you published an article, how much did they pay you?”

“Well, actually, I paid them.”

“What? So you just pay to publish?”

“Well, not quite, first it has to get past an editor and then a bunch of peer reviewers.”

“Oh, so that’s what the fees are paying for.”

“Well, actually peer reviewers are unpaid, and editors usually are too.”

“What?”

Academic publishing is a racket, and everybody knows it. Its profit margins are among the best in the business (nearly 40% for some large publishers). A lot of this depends on volunteer labour, which is getting increasingly scarce as the volume of article submissions ramps up (partly due to AI).

But the most important publication venue for many fields is arXiv (especially those related to computer science, physics, and math), the preprint server where research is free to publish and to read. Rather than conventional peer review, new authors generally need an endorsement to submit to a particular domain, and all papers are subject to basic screening. It went independent this year (it was previously housed in Cornell University) and hosts millions of papers on a budget of less than $7 million USD in 2025. At least in certain fields, arXiv is increasingly the terminal publication venue for many important papers, which receive extensive post-publication review and skip the traditional journal publishing model altogether.

Most academic publishing does not operate on this model of “free to publish, free to read”. Traditionally, journals are either open access or subscription-based. Commercial open access journals are free to read, but you have to pay (often extortionate) fees to publish. Subscription journals are free to publish in, but libraries have to pay (often extortionate) fees to read. But it’s common for even subscription journals to have some kind of fee to publish (“publication fees” or “article processing charges”) and it’s also common for subscription journals to offer “hybrid open access”, where you can pay open access fees to make that individual article free to read while the journal as a whole still requires a subscription.

Open access is obviously a very good idea in theory (research largely paid for by public money should be accessible to the public), which is why funders increasingly require projects they support to be published open access. But it’s also pretty clear that the big publishers have used this trend to wring even more money out of the academic publishing pipeline.

Anyway, a preprint server like arXiv is not a journal, since articles are not subject to traditional pre-publication review. But there are journals that operate on the same idea of “free to publish, free to read”. They are called “diamond open access” journals. This model is appealing, since journals mostly run on free labour anyway, and infrastructure costs can be pretty lean, as demonstrated by arXiv. Instead of financial support coming from authors or readers, it comes from academic institutions, professional societies, grants, and library consortia. Even then, the support necessary to sustain diamond open access can be unstable, and some journals previously operating under this model have transitioned away from it.

My impression is that most people in academia outside of the publishers themselves agree that the current model of publishing needs to change, as it relies on exploiting an increasingly stretched pool of free human labour to produce massive profits for a select few companies. Preprint servers like arXiv show you can have a massive research impact at a relatively modest cost.

There are also weird experiments like Unjournal that are trying to create alternatives to traditional publishing. Unjournal does not publish research; rather, it pays peer reviewers to rigorously review existing research and gives them proper visibility and credit for their work. Unjournal is a registered non-profit.

I guess my thought is, with all this AI billionaire philanthropy money that’s supposed to be coming down the pipe, how about setting some of it aside to support preprint servers and sustainable diamond open access journals? To help them build the prestige and credibility that is the only thing keeping the old business model of academic publishing alive?

I am tired of academia being in the thrall of a publishing oligopoly extracting exorbitant profits off of the free labour of academics everywhere.