ArXiv.org, the website that in 1991 pioneered the now widespread use of preprints as a fast, free alternative to scientific journal articles, will go it alone after more than 2 decades of being hosted by Cornell University. On 1 July, it will become an independent nonprofit corporation, according to an advertisement posted earlier this month seeking a leader for the new organization.
The move will help arXiv raise more money from a broader range of donors to fund the staffing and technology needed to support the site’s skyrocketing number of preprints—expected to top 300,000 this year—says Greg Morrisett, dean and vice provost of Cornell Tech, the graduate-education and research arm of the university that manages arXiv. The money will also help deal with other challenges, Morrisett says, such as the emergence of “AI slop,” mediocre or fraudulent submissions written entirely by artificial intelligence.
On social media, a few scientists have raised concerns about the shift, including that it might ultimately lead the platform to become a commercial enterprise. But Morrisett says users aren’t likely “to see any significant change in the near term. That’s our goal. Nothing’s broken about arXiv. We want to keep it successful. Hopefully they’ll see good things come out of this.”
ArXiv revolutionized scholarly communication in physics, its initial focus, as well as in other fields including computer science and math. In several subdisciplines, the majority of journal or conference papers appear first on the preprint server.
Recently arXiv’s growth has accelerated. Since 2022, it has expanded its staff to 27, in large part to deal with a 50% increase in submitted manuscripts. Besides AI-generated papers, the growth reflects a surge in the number of papers about AI, says Thomas Dietterich, a computer scientist at Oregon State University who chairs arXiv’s computer science editorial committee.
That growth has resulted in operating deficits in the past 2 years, including a $297,000 deficit in 2025. Cornell helped cover that overrun and also provided $819,000 in in-kind support, helping finance arXiv’s total annual operating costs of $6.7 million. Besides Cornell, arXiv is funded by more than 270 organizations, including academic library consortia and many of the world’s top research institutions, which pay as much as $10,000 per year in membership fees. (The fees are based on the volume of preprints posted by the member; the payments entitle them to a voice in arXiv’s governance and access to usage data.) Other major donors include the Simons Foundation and Schmidt Sciences, both of which recently helped fund technical upgrades that are nearing completion.
ArXiv competes with other funding needs within Cornell Tech, Morrisett says. “When we talked to certain organizations [about the preprint server], there was a sense of concern that Cornell exerted too much influence,” he says. “They would be a little worried about sending a check to Cornell University and hoping that it made its way into the arXiv coffers.”
He, arXiv leaders, and donors agreed that spinning it off was the best way to diversify and increase its funding. They had discussed the move for years, and it was recommended by Paul Ginsparg, the physicist who started the service when he worked at Los Alamos National Laboratory. When Ginsparg moved to Cornell in 2001, arXiv became part of Cornell’s library. (It moved to Cornell Tech in 2019.)
An academic institution proved not to be the ideal host for arXiv, Ginsparg says. “Universities, for all their virtues, have no history of providing this sort of global research infrastructure over a long term,” he says. Cornell, for example, had a limited capacity to pay software developers to maintain and upgrade the site, which still has a very no-frills look and feel. “This is not to say there won’t be significant short- and long-term challenges to getting this all to work” within an independent arXiv, Ginsparg says.
Cornell and Simons raised funds so the free-standing arXiv will begin life without a deficit and “enough funding [to operate] for a number of years,” Morrisett says. He adds that arXiv does not plan any immediate increase in membership fees.
In going independent, arXiv joins two other leading preprint servers whose creation it helped inspire: bioRxiv, which serves biological sciences, and medRxiv, which hosts preprints about medicine. Last year, they migrated from their original academic parent, Cold Spring Harbor Laboratory, to a new nonprofit, openRxiv, for similar reasons.
On social media, a few scientists have raised concerns about the proposed $300,000 salary for arXiv’s new CEO, saying it seemed high. (ArXiv disagrees.) Others worried funding pressures could eventually force the platform to be taken over by a for-profit company.
Ginsparg says he understands those fears. “But long-term sustainability has always been a concern,” he says. “If they think arXiv has been stable, they don’t know to what extent it was the Perils of Pauline all along”—a reference to a 1914 film melodrama—because the service has been run “on a shoestring.”
Ginsparg, who holds a faculty appointment at Cornell, continues to help run arXiv but says for years he has wanted to wind down his involvement and retire. He says that like the character Michael Corleone in The Godfather, he keeps getting pulled back in. He hasn’t decided what role he’d like to play in the new nonprofit. “I’ve been doing this for 35 years, and that’s roughly half my life. I want to get [the nonprofit] started, but at some point, obviously, you have to let go.”