Open by Default: How 2026 Became the Tipping Point for Open Access Publishing

For most of the twentieth century, the deal was simple: researchers produced knowledge, publishers packaged it, and libraries paid escalating prices to buy it back. In 2026, that deal has effectively collapsed. There was no single dramatic moment — just the quiet convergence of federal policy deadlines, funder mandates, maturing preprint culture, and a new generation of community-owned publishing infrastructure. Whether you are a graduate student submitting your first manuscript or a dean managing a library budget, the rules of scholarly publishing have changed. Here is what actually shifted, and what to do about it.

The Policy Dominoes Finally Fell

The biggest structural change comes from Washington. The 2022 White House Office of Science and Technology Policy memorandum — widely known as the Nelson memo — directed every US federal funding agency to eliminate the traditional twelve-month embargo on taxpayer-funded research. Agencies were required to have their policies fully in force by the end of 2025, which makes 2026 the first calendar year in which essentially all federally funded American research, from NIH clinical trials to NSF field ecology, must be publicly accessible the moment it is published. NIH moved early with its revised public access policy in mid-2025, and the rest of the federal research apparatus has followed.

The United States is late to a global movement rather than leading it. Europe’s Horizon Europe program has required immediate open access for years, cOAlition S funders have enforced Plan S since 2021, and UK Research and Innovation applies its open access policy to both articles and long-form publications. Latin America has operated large public platforms like SciELO and Redalyc for decades. What changed in 2026 is that open access stopped being a regional experiment or a funder preference. It is now the default expectation across most of the world’s major research systems, and the paywalled norm looks increasingly like the exception.

Preprints Became the First Draft of the Scientific Record

Physicists and mathematicians have shared preprints since the 1990s, and the pandemic years normalized the practice in biomedicine. In 2026, the preprint-first workflow has spread well beyond those early adopters. Psychologists, economists, climate scientists, and even scholars in parts of the humanities now routinely post working papers to servers like arXiv, bioRxiv, medRxiv, SSRN, and OSF Preprints before — sometimes long before — formal journal publication.

More interesting is what has grown up around the preprint. Services that review and curate preprints, from Peer Community In to journal-affiliated review models, have matured to the point where a reviewed preprint carries genuine weight in hiring, promotion, and grant evaluation. eLife’s decision to abandon traditional accept-or-reject decisions in favor of public reviews of preprints looked radical in 2023; in 2026 it looks prescient. Funders have followed suit, with major agencies explicitly accepting preprints as evidence of productivity in applications. The practical result: the version of record is no longer the only version that counts.

The Uncomfortable Truth: Open Is Not Free

Here is the tension defining scholarly publishing in 2026. Making research free to read does not make it free to produce, and the dominant commercial solution — charging authors article processing charges, or APCs — has created a new set of inequities. Average APCs at prestigious journals have climbed well past several thousand dollars, with some high-impact titles charging five figures. For well-funded labs at wealthy institutions, that is an irritation. For researchers in the Global South, at small colleges, or in chronically underfunded fields, it is a wall that replaced the old paywall with a publish-wall.

The interim fix, so-called transformative agreements between publishers and university consortia, is under strain. cOAlition S ended its financial support for these arrangements, arguing they were meant to be temporary bridges rather than permanent business models. Several large university systems have walked away from big-package journal deals entirely, choosing to fund open infrastructure instead of renting access from commercial publishers. The direction of travel is clear, but the transition is messy, and early-career researchers often absorb the friction.

Diamond Open Access Builds Real Infrastructure

The most encouraging development of the past two years is the rise of diamond open access — journals and platforms that are free to read and free to publish, typically owned by scholarly societies, universities, or community consortia. Long dismissed as a boutique ideal, diamond publishing now has serious institutional backing. The Global Diamond Open Access Alliance, launched in 2024, has coordinated funding and advocacy across dozens of countries. National open science plans in France and elsewhere channel public money into shared publishing infrastructure rather than APC budgets, and university presses are experiencing a quiet revival as institutions rediscover the value of owning their own outlets.

Diamond venues still face real challenges: sustainable funding, professional production quality, and the stubborn prestige economy that rewards journal brand names over publishing practices. But in 2026, for the first time, there is a credible, scalable alternative to both the paywalled and APC-driven models.

What Smart Researchers Are Doing Differently This Year

Policy shifts matter only if they change daily behavior. Researchers navigating 2026 well tend to share a few habits:

• Read the funder policy before choosing a journal. Public access requirements now shape where work can be submitted, not just where it ends up.
• Retain rights at submission. Rights-retention language, standard in Plan S jurisdictions, ensures you can deposit your accepted manuscript regardless of a journal’s embargo.
• Preprint early. Posting a preprint establishes priority, invites feedback, and satisfies most public access obligations immediately.
• Check waivers and institutional agreements before paying an APC. Many authors pay fees that a library agreement or waiver program would have covered.
• Deposit every accepted manuscript in a repository. Institutional and subject repositories remain the most reliable route to green open access.
• Plan for data and code sharing from day one. FAIR — findable, accessible, interoperable, reusable — data is increasingly a funder requirement, not a courtesy.
• Evaluate venues on practice, not just prestige. A journal’s openness, review quality, and community ownership are becoming legitimate selection criteria.

The Risks Nobody Should Ignore

None of this is risk-free. The APC gold rush continues to fuel predatory journals that mimic legitimate open access venues, and early-career researchers remain their primary targets. Market consolidation means a handful of commercial publishers still control an outsized share of the scholarly record and the analytics built on top of it. There is also a subtler danger: openness itself becoming another box-ticking metric, gamed and audited, while the deeper goal — knowledge as a genuine public good — gets lost in compliance paperwork.

The Real Question of 2026

The debate is no longer open versus closed. That argument is settled in policy, if not yet in every practice. The live question is what kind of open the academic world builds, and who pays for it. An open system financed entirely by author fees simply relocates the barrier from readers to writers. An open system built on shared infrastructure, community governance, and rights retention has the potential to be something better: a scholarly record that is genuinely public property. The researchers and institutions that treat openness as a design principle rather than a compliance chore are the ones shaping which of those futures wins — and in 2026, for the first time, the choice is genuinely theirs to make.