Where the sector is heading

The reduction in available funding sends signals that open science is no longer valuable enough to support monetarily. This philanthropic reshuffling is happening at the same time as US federal science funding cuts, leaving an ever-narrowing pool of funders responsible for critical infrastructure.

An infrastructure can be more widely used than ever and simultaneously more vulnerable to collapse. AI and other forces have contributed to skyrocketing traffic (and associated costs), while disintermediating users from content creators, often eroding certain monetization models. Funders, including companies that benefit directly from open infrastructure their models rely on, could move from ad hoc sponsorship toward proactive, sustained support for the infrastructure their own products depend on.

“Digital sovereignty” is no longer just a value, but a procurement and policy lever that could reshape which infrastructures get public legitimacy and funds in Europe. It sets a powerful precedent that funders and consortia could point to when arguing for large scale, sovereignty-aligned financing of research infrastructure. This move also illustrates increasing global mistrust of US-based tech and risks for US-based infrastructure providers, showing real consequences of US government and tech firms’ policies and behaviours. We are watching to see how digital sovereignty can manifest in regions that lack funding and capacity to fully disengage with US-based tech giants.⁷

In their rationale for narrowing OSF services, COS explicitly cites the ability to “lean on our open infrastructure colleagues and partners to fill use cases that OSF no longer supports.”⁸ COS’s work to proactively focus OSF’s future development is coupled with its encouragement of its users to embrace other open infrastructures for parts of lifecycle management that OSF no longer fulfils — displaying a clear sense of ecosystem responsibility and reciprocity. This decision bears resonance with Jisc’s decision to formally sunset IRUS, its institutional repository usage metrics service, in April 2026. Jisc’s Head of Product, John Kaye, attributed the decision to an evolving ecosystem and encouraged their former users to embrace a robust open alternative from OpenAIRE.⁹ Examples like these show how organizations are willingly giving up parts of their portfolio so they can focus on their unique contributions and acting in support of others. As open infrastructures look for sustainable solutions in a time of acute austerity, funders can encourage and directly support thoughtful cooperation.

Open infrastructure funders and maintainers have long argued for reuse as the end goal of openness. The PLOS report presents the case for open infrastructure as a facilitator of reuse, and illustrates the risk that, without it, the value of open science gets captured by vertically integrated commercial platforms instead. The report emphasizes the importance of coupling infrastructure improvements with a move away from the article processing charge (APC). These are collective action challenges that no single organization or sector can solve alone. Funders and institutions, in collaboration with publishers, can push for assessment reform, and the investment in the connective infrastructure this model depends on. The report explicitly highlights PLOS’ own continuing its investment in shared infrastructure like the Janeway platform.

HumanityAI represents one of the largest coordinated efforts to counterbalance corporate influence over AI’s trajectory with funding aimed at making sure people have a stake in the future of artificial intelligence. This kind of cross-sector alignment is rare in philanthropy, and its emergence here suggests funders see AI’s impact as too large and too cross-cutting for any single foundation’s portfolio to address alone. A coalition of this size and diversity can shape norms, set agendas, and provide opportunities for open infrastructures. Several of the initiative’s priority areas overlap with concerns central to open infrastructure work, particularly around who controls and benefits from the systems underpinning research and knowledge production. Connecting AI designed in the public interest with open science is a promising avenue for future support for open infrastructure, one that could help offset the evaporation of other, previously reliable funding sources.

A national government-backed program is treating infrastructure sustainability skills as a systemic gap worth funding directly, rather than leaving individual maintainers to figure out business models on their own. The work program explicitly acknowledges that technical infrastructure work and sustainability planning are distinct skill sets, and that the latter has been under-supported. Sustainability challenges have been brought into sharper focus by recent geopolitical crises that reveal the risks of depending on a single government or funder. While many funders require sustainability plans for infrastructure projects, the Dutch approach aims also to build the skills required for maintainers to craft such documented plans. Other countries and funders could look to this national-level action as a model. Rather than expecting each project or institution to develop this expertise independently, what would it look like to build in training and support for exploration of different business models, revenue diversification, and financial management, e.g., at the outset of a project rather than when it is in crisis?