Signals from the field

Introduction: Signals from the field

Because of the wide-ranging work we do at IOI, we often catch glimpses of emerging trends in open infrastructure and adjacent fields. In the 2024 report, we called this section Future Signals[1]; this year, we’ve adjusted the name slightly to reflect the fact that sometimes these signals help us to look back and reflect in order to see better ways forward. To that end, in this section now titled Signals from the Field, we explore three topics that have emerged repeatedly in our work this year.

These three pieces certainly tackle very different topics. The first piece is a meditation on observations related to the language used within our field, the second is an exploration of an adjacent field of digital public infrastructure, and the third covers our foray into researching measures of community health in open infrastructure. In compiling these pieces for this publication, we noted a particular throughline: a call to reflect on the values of open, whatever they might be in different areas of the field, and to articulate and express those values without reservations. In our first section on the binaries that shape our language around open, we explore how our words can be shortcuts for meanings that may become increasingly obscured. In the second section, we discover that the very concept of open is under debate, and tensions exist between the large-scale adoption of open technologies and the need to ensure those technologies serve their users. Finally, we consider some of the recent signals from the field regarding the health of open source and open infrastructure communities, and how those measures reflect the underlying needs of our communities for sustainable, values-aligned tools and services.

We hope you’ll enjoy reading these Signals from the Field.

Shorthand falls short: Why open infrastructure defies simple labels

In our work at Invest in Open Infrastructure (IOI), we often encounter assumptions that, on the surface, seem obvious or useful but quickly unravel under closer inspection. One of the most persistent of these is what we’ve come to call a binary fallacy: the assumption that complex, nuanced realities can be neatly divided into either/or categories.

This section explores some of these binary descriptions we’ve encountered in the open infrastructure ecosystem, particularly in our work on Infra Finder and through conversations across our network. Our aim is twofold: to illuminate these binaries for what they are — shortcuts — and to unpack the deeper meanings and assumptions that often accompany them and that have surfaced in important ways during our work this past year. To help with this section, we draw not only from our own experiences, but also from interviews with four colleagues who shared their perspectives on this matter: Geoffrey Bilder[2] and Joshua Neds-Fox[3] shared their expertise and observations from navigating the field, and Jennifer Roberts[4] and Katie Punia[5] provided their experiences working on and enacting some of these concepts in their time at Artefactual Systems, Inc.

We hope this reflection shows how language can both reveal and obscure the meanings and values encoded in open infrastructure work. We provide this initial reflection as a foundation for exploring two additional signals from the field that showcase the richness and complexity of open infrastructure work.

Beyond open vs. closed

Many readers may be familiar with one particular well-trodden example in the open movement: open vs. closed. This binary has long been used as a shorthand to describe the nature of access, participation, or licensing.

Many in the community have pointed out that openness exists on a spectrum. For example, the 2014 “How Open Is It?” guide for evaluating the openness of journals from SPARC maps out gradations of openness across six key aspects of journal publishing, reminding us that the term “open” is not a monolith. Guides like these and other spectra can help us describe what “open” means in a way that holds more value than a rigid definition. Furthermore, they can serve as aspirational frameworks to help guide development, instead of shutting out potential participants in our open infrastructure work. In this way, we view “open” as a gradable adjective, rather than an absolute one — something can be more or less “open” according to a set of descriptive parameters.

Following this assertion, we know that “open” can be expressed in various ways, but that there are some key shared values underlying the use of that term. Some of those are context-specific, such as the SPARC journal evaluation tool, which includes these parameters: reader rights, reuse rights, copyrights, author posting rights, automatic posting, and machine readability. Geoffrey Bilder shared a core value that underlies many of the context-specific parameters: that there is “an opportunity for users of a tool, service, or resource to exit”. If, for example, an open source software tool removes a feature that some of its users found valuable, those users should be able to go back to a previous version or fork the software to restore that feature. Suppose a researcher found a helpful article and the publisher removed it from their online archive; in that case, the researcher should be able to keep the copy they downloaded and continue to use it. If an organization provides a service to a group of members, those members should be able to retrieve their data and information in a useful way to move to another service.

Joshua Neds-Fox shared a related core value of “open” in his view: that a tool, service, or resource is open to “operating collaboratively with its community of users in a way that’s not focused simply on growth, extraction, or capture of the commons”. According to Neds-Fox’s view, this focus on community governance and input from users is a key mechanism to longevity and trust in open infrastructure. We see this as a complement to Bilder’s exit strategy assertion: open, community governance is a mechanism to avoid users’ need to exit, but the exit strategy can operate as a guarantee of real and meaningful community engagement.

As we see here, “open” can connote a wide field of meanings, including but not limited to access, participation, and licensing. At IOI, we see this first-hand with our entries in Infra Finder and the many different ways they express the various aspects of “open”. As such, we designed our eligibility criteria in a way that tries to capture some of these parameters, remaining flexible to try to capture what is most useful to our communities.[6]

The many ways of signalling values

One particular binary description has come up often since our last State of Open Infrastructure Report in 2024: commercial vs. non-profit. When we were designing Infra Finder initially, we used “commercial” and “non-profit” as basic business-type categories for infrastructure services. It seemed straightforward — until it wasn’t. Conversations with our global colleagues quickly surfaced the complexity within these broad terms. Some commercial vendors operate with deep commitments to mission and values. Some non-profits pursue market-based strategies and can lack transparency or community alignment. Furthermore, the term “non-profit” was not specific enough for some of our colleagues who desired more details to inform their understanding of how organizations operate. In response, we expanded our categories (to seven different general business types, and 20 different specific kinds of non-profit status, plus an “other” option).

As we did so, we started recognizing ways this binary is used in other contexts. Geoffrey Bilder shared an anecdote that aligns with our experience: at a conference, a new organization was slated to present a plan that promised to disrupt the regular operations of many others in the room. Everyone was tense, wondering about the motivations behind this new organization and its plan. The representative from the group began their presentation by saying, “We’re a non-profit.” Instantly, everyone in the room relaxed. The label triggered a sense of trust, community, and mission even before any further context. We’ve experienced the same when introducing IOI. Saying “non-profit” carries a cultural shorthand that suggests being mission-driven, people-centred, and values-aligned.

But as Bilder and others in our network have pointed out, this shorthand can be misleading. Some non-profit organizations don’t act in alignment with community values, and some commercial organizations do. That’s why watchdog organisations like Charity Navigator[7] and CharityWatch[8] exist; “non-profit” isn’t synonymous with “trustworthy.”

We wondered: what do we really mean when we say “non-profit” in this context? When our team talked to Joshua Neds-Fox for a recent project, he described conducting a “smell test” on organizations he was considering for partnership on a service. He aimed to determine “whether the organization showed signs of being driven by a mission and values” that aligned with his and his library’s, and “demonstrated a commitment to open governance and community contributions”. For this, an organization operating under a non-profit business model might be one factor that contributes to the signals of being a potentially good partner to work with for him and his organization, but it was not the only factor.

“Mission-driven” and “community-led” are other terms that we have seen used similarly as a signal of that alignment, but which also fail to capture the potential complexity. If the mission itself isn’t in harmony with the broader community or if that community is not the one leading the initiative — say, if a mission prioritizes the interests of shareholders above all else — that alignment may be counterproductive in a values-based ecosystem like open infrastructure. We see this prioritization of values in documents like the World-Historical Gazetteer’s Digital Initiative Sustainability Report (Mostern & Straub, 2025), which lays out the need for open infrastructure organizations not only to develop, but frequently return to, “a clearly articulated and consistent set of values” and showcases this practice across several case studies of digital humanities and allied initiatives.

Innovation and maintenance

Another binary that has surfaced repeatedly in the past year is a distinction between “research” and “infrastructure” and, relatedly, between “creators” and “maintainers.” These terms and their related constellations of meaning shape funding landscapes, perceived value, and even career trajectories.

Much of the funding in our ecosystem is geared toward innovation: pilot projects, new tools, and experimental platforms. We see this in our funding analysis in the earlier section of this report entitled “The state of open infrastructure grant funding.” This new development work is certainly exciting and vital. But what happens once those platforms and experiments are completed and the final report is sent to the funder? Maintenance — the ongoing work of keeping those tools running, updating documentation, and fixing bugs — is too often underfunded or invisible, and leads to a deprecation of projects and a long-term sustainability issue.

This isn’t a new problem. It’s well-documented that maintainers are overburdened, underappreciated, and ageing out (see, e.g. Nagle et al., 2020; Tidelift, 2024). This is in tension with the need expressed in our community for proof of longevity when selecting infrastructure. Joshua Neds-Fox shared that when he is choosing an infrastructure to partner with, he thinks about a system’s utility decades down the line, and whether it will continue to evolve with its community of users or if it will be captured by a corporate entity and incorporated into proprietary structures. Open infrastructure’s long-term sustainability and longevity can establish trust in this space. That longevity doesn’t happen without maintenance, and maintenance suffers without funding and attention.

So why does the “creator” or “innovator” role get more attention and reward in the funding environment? There is a lot of semantic baggage tied up in these words, and funding mechanisms often define sustainability narrowly as something that can outlive a grant cycle without ongoing financial support. But what if we described sustainability instead as more aligned with maintenance— as a network of relationships, practices, and shared commitments that evolve over time? We explored this concept in a case study through a conversation with Jennifer Roberts and Katie Punia from Artefactual Systems, Inc.

A case study from the field: Artefactual Systems Inc.

We first met the team from Artefactual Systems, Inc., (hereafter referred to as Artefactual) when we recruited entries for Infra Finder from their open source digital archiving tools Archivematica and AToM. Artefactual was founded by Peter Van Garderen in 2001 and is incorporated as a private for-profit company in Canada. The goal of Artefactual was to create alternatives to commercial digital preservation software that were both technically robust and financially accessible for archivists working in a variety of organizations and conditions. This vision materialized in a model where the software developed by the group is free and open source, and they generate revenue through professional services such as consulting, support, training, and development (Van Garderen, 2009). The software and services are all directed under Artefactual’s four values: openness, collaboration, sustainability, and trustworthiness.

We asked about Artefactual’s business model and how it helps them do the work in their space. As a private, for-profit company, Artefactual is not externally funded, and the six shareholders are all current or former employees who work within the company’s shared values. This business model allows the company to both provide open source tools that are valuable in the digital preservation field, and provide fair compensation and benefits for the people who work for the company. Their operations push against the idea that “Free and Open Source Software” must be free in every aspect. Jennifer Roberts, Systems Archivist, summarized their approach to their business: “The code is free. The ability to use it is free. Our time isn’t free.” This clarity helps manage expectations around community support while reinforcing the importance of sustainable labour practices in the open ecosystem.

One example of an Artefactual initiative that employees see as aligning practice with values is the creation of “contributor gatherings”, designed to facilitate collaboration between the company’s maintainers and active members of the user community.[9] These meetings help build continuity in development, foster a distributed support ecosystem, and ensure that the software’s ongoing development responds to the community’s needs. Jennifer Roberts said about this work that, “Colleagues describe us as a node on a network. Rather than controlling everything about the project, it’s about distributing the work, understanding the community’s needs and practices, and trying to facilitate a way to make sure our tools will last into the future.” At the same time, they acknowledge the reality for many software projects that there will come a time when features and entire software packages may become obsolete. Echoing Geoffrey Bilder’s comments in the previous section, Roberts noted that they take a data-focused approach where users can easily take their metadata and digital objects, or use the connectivity built into the tools to send their data elsewhere. This “exit strategy” and intentional counter to the possibility of lock-in is key to how Artefactual operates as an organization, and maintaining those connectors and dependencies has been a focus of their recent development work, sometimes in tension with requests from the community for new features. According to Roberts and her colleague Katie Punia, this intentional prioritization of foundational upkeep over innovation reflects the company values of sustainability and trustworthiness, as well as their understanding of digital preservation as a “practice of care” for the material being preserved, the users who depend on the software, and for the developers doing the work (Simpson, 2020).

This short case study shows one organization’s approach to operating beyond the binary distinctions of commercial vs. non-profit and open vs. closed. It shows one way that an organization can show alignment and shared values with its community of users beyond presuppositions encoded in company types.

Looking forward: Language to empower infrastructure

The message of this section is not intended to call out specific uses of language. Binaries can be tempting; they’re easy, fast, and sometimes rhetorically powerful. They can be useful shorthand in various situations, yet they also run the risk of flattening complexity and obscuring nuance. And when we’re not all on the same page — or when we presume we are but don’t articulate our presuppositions — these binary distinctions can carry unintended meanings that erase valuable information and shape decisions in ways we don’t yet fully understand.

At IOI, we believe that language is part of the infrastructure. The words we use to describe our work shape how it’s understood, valued, and supported. So when we talk about being open, sustainable, or values-aligned, we want to be precise and generous with our descriptions.

We hope that you will keep this discussion in mind when reading the next two sections of our Signals from the Field. When looking into the alignment between digital public infrastructure (DPI) and open infrastructure, for example, we discovered a basic disagreement about what “open” means in the context of DPI and in articulating the goals and values of DPI work. And when it comes to articulating goals and values, this approach is embedded in some community health frameworks such as the Principles of Open Scholarly Infrastructure, CHAOSS Metrics, and the FOREST framework; even though uptake of these frameworks has been limited to specific communities, some of their components are consistently on the minds of the people who work in open infrastructure.

References

Collister, L. (2024, November 20). Describing open. Katina Magazine. https://doi.org/10.1146/katina-112024-1

Mostern, R., & Straub, A. (2025, February). Digital initiative sustainability report. University of Pittsburgh: World Historical Gazetteer. https://heyzine.com/flip-book/270892f95d.html

Nagle, F., Wheeler, D. A., Lifshitz-Assaf, H., Ham, H., & Hoffman, J. L. (2020, December 10). Report on the 2020 FOSS contributor survey. The Linux Foundation & The Laboratory for Innovation Science at Harvard. https://www.linuxfoundation.org/resources/publications/foss- contributor-2020?hsLang=en

Simpson, J. (2020, November 9). The business of preservation is the practice of care. Digital Preservation Coalition Blog. https://www.dpconline.org/blog/wdpd/blog- justin-simpson-wdpd

SPARC, PLOS, & OASPA. (2014). HowOpenIsIt? A guide for evaluating the openness of journals. https://sparcopen.org/our-work/howopenisit/

Thaney, K. (2025, January 15). What we talk about when we talk about open infrastructure. Katina Magazine. https://doi.org/10.1146/katina-01152025-1

Tidelift. (2024, September). The 2024 Tidelift state of the open source maintainer report. https://tidelift.com/open- source-maintainer-survey-2024/

Van Garderen, P. (2009). The ICA-AtoM project and technology. Association of Brazilian Archivists. https://ica-atom.org/download/VanGarderen_TheICA-  AtoMProjectAndTechnology_AAB_RioDeJaniero_16- 17March2009.pdf

1. https://investinopen.org/state-of-open-infrastructure-2024/sooi-future-signals-2024/
2. https://gbilder.com/
3. https://library.wayne.edu/info/staff-directory/dp5745
4. https://www.artefactual.com/our-team/jennifer-roberts
5. https://www.artefactual.com/our-team/katie-punia
6. https://investinopen.org/add-an-infrastructure-to-infra-finder/
7. https://www.charitynavigator.org/
8. https://www.charitywatch.org/
9. https://www.artefactual.com/post/gathering-the-contributor-community

The intersection between digital public infrastructure and open infrastructure

Over the past 10 years, the concept of digital public infrastructure (DPI) has gained traction globally. As a field, DPI is still evolving and as a result, many things about how to implement, govern, and contextualize DPI are still in flux. What is clear though is that DPI and open infrastructure are fields that already have a lot in common and could even in future both learn from each other. Some of the tensions in DPI, ranging from issues with the definition of the concept of sustainability, to the role of community in governance, are also all challenges that open infrastructures struggle with, even to date.

For this chapter, we spoke to two experts in the field, Mila Samdub, Visiting Fellow at the Yale Information Society Project studying digital government infrastructure in India and Emmanuel Oloo-Khisa, Africa Director at the Centre for Digital Public Infrastructure (CDPI). They shared their perspectives on some of the interplays between DPI and open infrastructure. This article summarizes the key thematic areas that surfaced through our conversations.

Demystifying DPI

Digital public infrastructure is a concept that has gained traction over the past decade or so, but what does it really mean? DPI can be broadly defined as platforms that facilitate the provision of digital products and services, built to enable access to a range of “essential society-wide functions,” such as services in the areas of identification, payments, and data exchange.

Two other key terms are also used interchangeably in the DPI space. According to the Digital Public Goods Alliance,[1] Digital Public Goods (DPG) refers to technologies such as open-source software, open standards, open data, open AI systems, and open content collections that adhere to privacy and other applicable laws and best practices, do no harm, and help attain the United Nations (UN) Sustainable Development Goals (SDGs).[2] The Digital Public Goods Alliance maintains a registry of DPGs, vetting them against baseline requirements that must be met and fulfilled to earn recognition as a DPG. In theory, DPGs provide the open-source building blocks for DPI, allowing governments and communities to adopt interoperable tools without vendor lock-in.

Public Digital Infrastructure (PDI) is a relatively new concept that refers to digital infrastructures designed to maximize public value by combining public attributes with public functions and various forms of public ownership (Krewer & Warso, 2024). Unlike DPI, which can lean heavily on centralized, state-led models, PDI often reflects a more grassroots, decentralized approach.

While DPI enables access to and transfer of information, and often relies on open standards and interoperability between systems, it is not necessarily built on open source solutions. Payment apps like Unified Payments Interface (UPI) in India and Pix in Brazil use proprietary technology and operate as public-private partnerships. These types of partnerships can bring needed capacity, resources, and expertise to bear on societal needs, but can also present ethical and logistical dilemmas. In India, for example, Samdub notes that public-private partnerships may circumvent government guidelines that recommend the use of open source software. Industry involvement in the public sector also provokes questions about how “DPIs are using public data to support the private capture of that value.”[3]

The conflicting definitions of “open” in DPI are similar to the debates in the open infrastructure space, where there have been numerous discussions about defining open infrastructures. Some have tried to define open infrastructures holistically, and others lean toward describing open infrastructures based on specific characteristics (SCOSS, n.d.; Collister, 2024).

DPI principles and practice

Some key concepts distinguishing DPI from other systems are human centred design philosophy, robust governance, and market/private sector participation (Sang et al., 2025).

Within the human-centered design philosophy there are several key criteria that DPI has to incorporate. One is that DPI needs to be people-centric to ensure that the public’s rights to privacy and autonomy of their data are respected. A key element of the common design philosophy is that DPI needs to be interoperable to allow the sharing of data across different systems. Closely related to this interoperability is the fact that DPI needs to be modular (extensible) to allow components to be built on top of each other over time. Another design philosophy is that DPI in principle favours data federation (sharing data in a secure and controlled manner over different components) over data centralization. The final design philosophy is that DPI should be protocol based — using open protocols and standards to allow for diverse innovation within the system, rather than relying on a single platform.

Another key feature of DPI is participatory governance. This is critical because, given the sensitive data domiciled in DPI systems, it is essential that no one singular entity gains undue control over them. These safeguards include approaches like Privacy by Design (where security is built in the code) or Privacy by Policy (where after the infrastructure is built, there are intentional policy frameworks implemented to enhance data security).

Market/private sector participation is meant to ensure that the private sector is an active participant in the increased adoption and scaling of DPI, that there is enough latitude to ensure a competitive market that allows businesses of all sizes to participate, and that incentives are provided to avoid market capture by powerful players.

Oloo-Khisa commented, “Other tech concepts differ here because, for the first time, we are mainstreaming the whole idea of safeguards and then market play. DPI is a technology for the good of society but with a clear thought process around market enablement.”

Samdub commented, “While several of the principles invoked in relation to DPI are admirable, in practice several DPI deployments, including model examples like India’s Aadhaar identity system, don’t follow these principles. Aadhaar, for example, centralizes data storage, leading to date security concerns. Its development excluded civil society participation. And it exhibits several characteristics of a platform, leading some scholars to argue that it has become a form of alt big tech (see Parsheera, 2024).” 

Contextualizing DPI in developing economies

In the formative stages of DPI, a number of non-profit organizations and philanthropies were at the forefront of promoting DPI use globally. While, in principle, this is a good thing, it poses a challenge for countries to apply DPI infrastructures that were built in different countries with different realities and, therefore, not aligned with the needs of local contexts. One example of exported DPI is Modular Open Source Identity Platform (MOSIP), which was built off of India’s Aadhar system, where 17 out of the 26 installations of the system globally are in Africa (Santhanam et al., n.d.). The widespread adoption of MOSIP can be attributed to its interoperability, cost effectiveness but also a lack of African alternatives.

“One way DPI can be more aligned to the local realities in the developing economies is by setting up supporting ecosystems in the creator and developer economies,” said Oloo-Khisa. “There is also a need to work on strengthening governance via platforms like governance roundtables or circles and increasing the amount of funding allocated to support the implementation and maintenance of DPI. Equally important is developing a belief or a can-do attitude that developing nations can build their own DPI systems and not just buy and customize from other contexts.”

Some of the ground realities that make developing DPI in developing economies challenging are as follows:

• An inherent friction between need and speed of rollout: the need for the DPI is urgent for people, but there are often challenges in speed of implementation;
• The local capacity to architect and build minimalist technology infrastructure solutions may be low or vary at different levels of government;
• There may be funding or budget gaps for execution (not just for the new software systems but also for compute/hardware); and
• Procurement cycles may be long and tedious to kickstart progress. There are multiple hoops to jump through before a country can even test out DPI through a pilot.

One of the parallels that can be drawn between DPI and open infrastructure is the fact that development in both sectors is undergoing rapid change. In open infrastructures, it is commonplace that in developing economies, practitioners get tools that have been developed elsewhere and customize them to their own needs. In the DPI space, similar patterns are evident.

Centralization vs. decentralization: Who builds and who governs?

One of the central tensions in digital public infrastructure today is between centralized, state-led deployment and decentralized, community-driven initiatives. Over the years, DPI implemented at scale has often emphasized adoption metrics (number of registered users and number of people using a particular system) as an indicator of success (Dolan & Satapathy, 2024). To achieve rapid, large-scale adoption, governments may opt for centralized, top-down models due to reduced complexity in interoperability across ministries, consistency in service delivery, and control over implementation.

However, this scale and speed can come at a cost: less transparency, reduced flexibility, and minimal space for civil society or grassroots actors to shape the system.

As Mila Samdub observes, “You get a bias from building out a system as fast and as big as possible... it’s not clear that’s the best way to deliver public value.”

By contrast, decentralized or federated approaches — where local or regional actors co-create and govern digital systems — may better reflect community needs and values. These models align more naturally with open-source principles like transparency and collective ownership. Yet they face serious challenges: coordination, technical capacity, and lack of funding can all impede implementation at scale. Currently, the dominant approach is the centralized model. Still, given the increasing awareness of the need for more localized control of people’s data, decentralized DPIs could gain ground in the future.

“Could you have a digital public infrastructure that has ownership and governance at a more local or regional scale?” Samdub asks. “These are questions we need to be thinking about.”

Balancing public and private interests

In DPI, one of the important aspects is the partnership between the private and public sectors to facilitate the provision of essential services to the citizenry. The public and private sectors have a symbiotic relationship — governments need the capital investment and technical expertise of the private sector, which in turn needs the government to provide them with a conducive business environment.

Transparent and participatory governance is one key pillar of the DPI approach and is crucial in ensuring that dominant players do not gain undue influence over DPI implementation at the expense of the common good.

“The questions around forking of code, source code, and how we contract for DPI are evolving questions. CDPI attempted to address some of this by creating what we call DPI as a package Solution (DaaS) which refers to the rapid deployment of DPI, through upgrades of existing infrastructure (instead of greenfield implementations) through non-procurement routes.[4] In this case, DPI software developers can build a DPI and then make it readily deployable on popular cloud solutions such as Google Cloud or Amazon Web Services, while a separate institution manages the source code, contracting, hosting of different versions, upgrades, and maintenance. There are several instances of this dual-structure model — one that separates community-based development from institutional-level oversight and management. Some of the organizations that have deployed this approach are Red Hat,[5] Drupal,[6] and OpenCVRS.[7] DaaS is a more straightforward way to implement DPIs and can help circumvent some of the complexities in DPI implementation that we are already seeing in the market now. For example, with the development and implementation of e-Citizen in Kenya, where details on the ownership of the platform and how data is utilized has caused some controversy,” remarked Oloo-Khisa. (For more on e-Citizen, see Daily Nation Editorial Board, 2025.)

We can also see a corollary between DPI and open infrastructures in the tension between public and private interests (Samdub, 2025). In open infrastructures, there is a big tension when publicly funded research is then domiciled in repositories and journals owned by private entities.

Innovations and future outlook for DPI

DPI, as with many other technologies, is rapidly evolving. Advances in blockchain and artificial intelligence, the need to ensure alignment with community needs and decentralization of DPI are some of the emergent topics that may influence how DPI evolves in the future. We asked Samdub and Oloo-Khisa to share what they are excited about in the DPI space going forward.

“There are things that I find inspiring that I wish would be part of the DPI conversation,” said Samdub. “Community networks are forms of commons-based ownership and sovereignty over connectivity. I wish there could be more conversations between community networks and DPI, which has unfortunately largely ignored the question of connectivity. In India, we have Social Audits, a transparency mechanism, especially in public service delivery, which are built in by law to certain programs. This is a collaborative process where the public creates an open, horizontal space to hold the government accountable. Both these community-centric approaches are characterized by a sort of popular sovereignty and accountability that’s entirely missing from DPI as we know it.”

On the other hand, Oloo-Khisa is interested in developing the capacities and a conducive environment for DPI to thrive in developing economies. “As DPI uptake in Africa continues, there are a number of issues that we are trying to address like the shortage of developers, creation of an enabling environment for DPI, and also mobilizing resourcing to support DPI implementation and adoption,” said Oloo-Khisa. “The Africa Digital Economy Lab (ADEL) is a virtual secretariat at the moment that aims at building an environment that facilitates policy conversation on DPI, aggregating funds to support development of the ecosystem, and finally, creating a support structure that helps countries to implement via developing much needed skills like project manager, project quality assurance, and project technical architects that help countries develop DPI solutions.”

Conclusion

This section was meant to provide an initial exposition of the intersection between two fields that are gathering traction globally. DPI and open infrastructures share a lot in common ranging from how in principle they approach issues like responsible data management, to governance and participation, among others. However, there are also issues that both DPI and open infrastructures are grappling with like how to ensure the interest of the public is always centered despite government and private sector interests, sustainability, as well as impact assessment. While this section provides valuable foundational context, there are still plenty of additional insights that we hope to highlight in the future.

References

Center for Digital Public Infrastructure. (n.d.). DaaS in a nutshell. DaaS in a Nutshell. https://docs.cdpi.dev/initiatives/ dpi-as-a-packaged-solution-daas/daas-in-a-nutshell

Collister, L. (2024, November 20). Describing Open. Describing Open. https://katinamagazine.org/content/article/ open-knowledge/2024/describing-open

Daily Nation Editorial Board. (2025, March 10). Private ownership of eCitizen a grave risk. Daily Nation. https:// nation.africa/kenya/blogs-opinion/editorials/private- ownership-of-ecitizen-a-grave-risk-4959540

Desai et al., n.d.. How digital public infrastructure supports empowerment, inclusion, and resilience. https://blogs. wrldbank.org/en/digital-development/how-digital-public- infrastructure-supports-empowerment-inclusion-and- resilience

Digital Public Goods Alliance. (n.d.). Digital Public Goods. Digital Public Goods Alliance. https://www.digitalpublicgoods.net/digital-public-goods

Dolan, J., & Satapathy, S. (2024, March 27). As investment in DPI increases, how can we meaningfully measure its impact? https://dial.global/how-to-measure-dpi-impact/

Krewer, J., & Warso, Z. (2024). Digital Commons as Providers of Public Digital Infrastructure. Open Future. https:// openfuture.eu/wp-content/uploads/2024/11/241113_Digital- Commons-as-Providers-of-Public-Digital-Infrastructures.pdf

Lim, Q. (2024, February 1). NEXT-GENERATION DIGITAL PUBLIC INFRASTRUCTURE. https://www.elevandi.io/hubfs/ Next%20Generation%20Digital%20Public%20Infrastructure. pdf

Parsheera, S. (2024, July 10). Digital Public Infrastructure and the Jeopardy of “Alt Big Tech” in India. Digital Public Infrastructure and the Jeopardy of “Alt Big Tech” in India. https://casi.sas.upenn.edu/iit/smriti-parsheera-2024

Samdub, M. (2025, March 14). What Would it Take for Digital Public Infrastructures to Actually Serve the Public? The Wire. https://thewire.in/tech/what-would-it-take-for-digital-public- infrastructures-to-actually-serve-the-public

Sang, D., Sambuli, N., & Munga, J. (2025). Digital Public Infrastructure: A Practical Approach for Africa.

Carnegie Endowment for International Peace. https:// carnegieendowment.org/research/2025/02/digital-public-  infrastructure-a-practical-approach-for-africa?lang=en

Santhanam, N., Narayanan, R., & Rajagopalan, S. (n.d.). Journey Towards a Billion IDs. Modular Open Source Identity Platform (MOSIP). https://connect.mosip.io/resources_ pdf_2025/Journey%20Towards%20a%20Billion%20IDs.pdf

SCOSS. (n.d.). Defining open infrastructures. https://scoss. org/what-is-scoss/defining-open-infrastructure/

1. https://www.digitalpublicgoods.net/
2. https://www.digitalpublicgoods.net/digital-public-goods
3. https://www.internetgovernance.org/research/india-stack-public-private-roads-to-data-sovereignty/
4. https://docs.cdpi.dev/initiatives/dpi-as-a-packaged-solution-daas/daas-in-a-nutshell
5. https://www.redhat.com/en
6. https://www.drupal.org/association
7. https://www.opencrvs.org/

Trust, transparency, and technology: Do community health frameworks shape open infrastructure decisions?

Open infrastructure is a socio-technical construct that needs people, culture, technology, and funding to survive and thrive. The collaborative approach that addresses these needs to ensure the ongoing success and sustainability of the open infrastructure is called “community health”.[1] In recent years, various frameworks — collections of principles, ideals, and metrics — have been designed to help foster understanding of the community health of open infrastructure in the research ecosystem. These frameworks are useful as a tool for orienting people and organizations toward a collective intention, particularly when they are reinforced through incentives and accountability mechanisms.

With the support of a grant from the Digital Infrastructure Insights Fund, Invest in Open Infrastructure (IOI) conducted discovery interviews with a cross-section of members of the open infrastructure ecosystem, including academic libraries, research institutions, funders, technology creators, vendors, and collectives.[2] We wanted to understand what impact, if any, community health frameworks have had on the use, operation, and funding of open infrastructure. To that end, we sought to interview people from organizations that were not signatories to the frameworks in order to get a clearer sense of overall priorities and how widely known the frameworks are.

Overview of community health frameworks

In our interviews, we focused on three community health assessments that have garnered attention within the open research infrastructure ecosystem:

1. The Principles of Open Scholarly Infrastructure (POSI)[3] — Introduced in 2015, POSI is designed to “guide the operation and maintenance of open scholarly infrastructure” through 16 principles that cover governance, financial sustainability, and “insurance” that code, data, and patents will not lock-in users.
2. Community Health Analytics in Open Source Software (CHAOSS) Metrics[4] — Launched in 2019, this framework focuses on open source software and its contributors, communities, companies, and foundations, helping them understand the health of the open source projects they engage with.
3. The FOREST Framework for Values-Driven Scholarly Communication[5] — Released in 2022 by the Next Generation Library Publishing project, FOREST is intended to help scholarly communication organizations and communities demonstrate, evaluate, and ultimately improve their alignment with values including: financial and organizational sustainability, openness, representative governance; equity, accessibility, and anti-oppression; sharing of knowledge, and transparency.

This section contains preliminary findings and a sample of the themes we’ve observed so far. Additional reporting on this project will be released following this report’s completion in late 2025.

Framework awareness and underlying values

A notable finding from our interviews was that awareness and use of these frameworks was generally low.

Most interviewees were unfamiliar with POSI, CHAOSS, or FOREST, or had encountered them only in passing. Despite this, when asked about how they view a healthy open infrastructure community, participants consistently emphasized priorities such as transparency, sustainability, and community governance, aligning closely with the frameworks’ goals. This finding suggests that the values embedded in these frameworks are present in the ecosystem, even if the specific framework tools themselves are not widely referenced.

Another notable finding is that community health is rarely the deciding factor when organizations select infrastructures to use. Participants described decisions about whether to adopt or support infrastructure tools as driven primarily by practical needs: does the tool do what we need it to, and does it meet our budget constraints? In some cases, interviewees also wanted to find out who else was using the infrastructure to better understand if it was an appropriate selection. This basic question of “fit” for the needs of an organization indicates to us that open infrastructures and vendors could improve their chances of being selected if they prioritize making it easy to understand what features and interoperability they offer, and list of who is using their solution, or what their ideal users look like.

While most decisions were based on practical considerations such as cost and fit for purpose, community health attributes were seen as value-added factors which could contribute to comfort and trust in the decision.

Relationships matter: The many facets of trust

Given the community-and-collective nature of open infrastructure, trust plays an outsized role. Interviewees described how difficult it is to know what’s best-suited for any single organization’s needs, constraints, and intentions. Interviewees shared that gathering as much relevant information as possible was crucial to their decision making.

Most frequently, they talked about how important it was to gather information by talking with people involved in the infrastructure. They named that it was also useful to review public-facing communication such as engagement messaging, technical discussions, feature lists and user testimonials to get a sense of technical and social-cultural fit. However, these same interviewees shared that when it came to assessments, the information presented might not, and in some cases can’t, expose uncomfortable realities.

Some interviewees said they would be more comfortable with their final choices if they could easily know about more alternatives and have deeper trust in solutions or partnerships, but it was difficult to find information and develop sufficient interpersonal trust ahead of decision time. We see an opportunity for OIs to share more information about their user base and capabilities, conducting outreach/engagement to build awareness and trust amongst existing and new audiences.

Some organizations had the capacity and organizational desire to contribute labour or funds back to the community, or saw community governance as a proxy promise to prioritize the institution’s needs. Members of these organizations said that governance, community contribution, and financial transparency heavily factored into their evaluations. With this intention, interviewees repeatedly emphasized the importance of trust in their decision-making processes. They wanted to know who maintains and governs infrastructure projects, how finances are handled, and under what circumstances labour contributions are integrated.

Interviewees who were, or who represented, OI users wanted to know how welcome they would be in the community, how much influence they would actually have, how long and by what mechanisms would they gain influence, and whether there were political or personality frictions that might cause difficulties.

For those who intended to fund an OI, it was important to be in touch with people responsible for finances, leadership, and work because readily available an accurate financial data was hard to come by, and funders wanted confidence that financial contributions would advance their larger goals.

Apprehension about framework use

Among the few participants who were familiar with community health frameworks, most had not applied them in practice. We heard a theme of concern about exposing shortcomings which could result in negative outcomes. As one participant put it, very frankly: ”No one likes being judged.”

We also heard a contrasting perspective: that transparency about shortcomings could be an opportunity for values alignment, which could both foster trust and create opportunities to offer support.

Participants who had considered using assessment frameworks said they felt more comfortable using them as a self-assessment to help improve their own organization or project. There was also a general acknowledgement that a third-party evaluation would carry more weight, but would require prioritizing evaluation above other community needs and desires.

Fear of negative consequences surfaced specifically when funders encouraged the use of community health frameworks as part of grant activities, even when these activities were framed as growth and development activities rather than requirements or criteria for qualifying for grants. An interviewee from a funding organization noted: “We’re used to measurement being associated with competition. We’re not accustomed to measuring in service of cooperative behaviors.”

A pattern we saw in the interviews was that organizations needed or wanted to reach a certain level of organizational and operational maturity before even considering a self-assessment. One interviewee noted: “Health frameworks are useful but often feel aspirational. They should be more explicitly tied to solving real problems.”

Additionally, interpretation of key framework values varied widely. For example, the idea of “community governance” meant different things in grassroots, collective-led groups than it did in institutions with hierarchical decision-making structures. Some interviewees said they modified or included individual framework bullets as guidelines or hard criteria in their own documents or evaluation rubrics. POSI and the It Takes a Village Guidebook[6] were the most commonly referenced in these modified uses.

Moving forward

Despite a general lack of familiarity with the community health frameworks, there is a shared desire across the ecosystem for clearer signals of functional capabilities, transparency, collective trust, and alignment from open infrastructure projects. Interviewees expressed that better communication, more visibility into how the organization is governed, who its users are, and what opportunities exist to support capacity building would all help infrastructure users make more confident, values-aligned decisions. OIs themselves expressed appreciation for the aspirational nature of the frameworks, and the opportunity to use them for development of their organizations as they mature.

As we finalize this phase of our research, we welcome your insights and ideas for how to make it easier to select, contribute labour to, or fund open infrastructures, whether by making improvements to Infra Finder or by some other means.

If you have thoughts or suggestions, please send them to us at research@investinopen.org.

1. Our initial analysis revealed confusion between “community health” as it relates to open source software and its meaning in public health. We’ll further clarify this distinction in future reports.
2. https://infrastructureinsights.fund/
3. https://openscholarlyinfrastructure.org
4. https://chaoss.community/kb-metrics-and-metrics-models/
5. https://doi.org/10.5281/zenodo.6557302
6. https://itav.lyrasis.org/

Feedback