2025 State of Open Infrastructure Report

"At IOI, we have long wrestled with what makes an open tool, protocol, platform, system, or network an “open infrastructure.” The term is particularly important for us; not only is it in our name, it also resides at the core of our mission to increase investment in and adoption of open infrastructure.

While we have come to no single, let alone strict, definition of the term, we have often returned to a key element, a kind of “true north” that stands out as we try to differentiate between what is and is not open infrastructure: reliance.

At IOI, we are focused on the research ecosystem, and we take “reliance” seriously. Researchers, curators, and technologists rely on open infrastructure as critical underpinnings of our information landscape. Disruption in the availability of open infrastructures in the research ecosystem can jeopardize swaths of human knowledge.

Imagine just a few of the major open infrastructures serving scholarship today: arXiv, PubMed, DSpace, Open Journal Systems, or perhaps Creative Commons. Consider how much of our memory, our knowledge trail, depends on open research infrastructures, and consider how many of these have been created on a cycle of grants and sustained on a combination of donations, membership fees, service hosting, and/or specialized development. Repositories and publishing platforms, metadata standards and persistent identifiers — as long as they all work reasonably well, most scholars, teachers, and researchers across disciplines and focal areas accept them as a given and depend on them as a reality. But these infrastructures are not, on the whole, sustainable businesses with robust fiscal models and diversified revenues. Many do not turn profits; most operate at steady losses that are absorbed by philanthropic and government funders and a variety of research institutions including labs, universities, and colleges. They operate on systems rife with technical debt, and they depend upon volunteer labour to cover much of their human costs from governance to editorial review to code development...."