The Soft Infrastructure We Forgot: Why Digital and AI Efforts Fail Without Leadership Systems
DARO collaborated on this article with Alethea Hannemann, the Co-Founder and CEO of Board.Dev. Alethea’s built cross-sector collaborations at Taproot, Okta, Google, and Splunk. She’s the former COO of a $25M nonprofit and co-author of Powered by Pro Bono. She bridges tech, philanthropy, and mission leadership.
The hidden reason digital projects fail
Across the social sector, digital and AI projects fail in remarkably similar ways. Organizations launch new CRMs, dashboards, and AI pilots, only to watch usage drop and systems drift out of sync with reality. Momentum fades, frustration grows, and what once felt promising becomes another abandoned tool.
These failures look technical, but they are actually structural.
Most organizations have invested heavily in the hard infrastructure of data and technology—platforms, tools, data warehouses, AI models—but not the soft infrastructure required to make those systems useful, meaningful, and sustainable.
These failures are the predictable outcome of how philanthropy funds digital and data work: time-bound, tool-centric, and focused on delivery rather than stewardship. Across regions and issue areas, funders underwrite platforms and pilots while leaving leadership systems like governance, decision rights, and learning routines, unfunded and unnamed.
These leadership systems are rarely budgeted or assigned owners, yet they are what determine whether digital systems thrive or collapse.
Digital and AI failures are fundamentally leadership failures rooted in neglected soft infrastructure, and AI raises the stakes. Weak soft infrastructure that once slowed progress now creates real risk, automating flawed assumptions, embedding inequity, or creating tools that no one can maintain.
What soft infrastructure is, and why it determines success
Hard infrastructure is what technical systems get built and soft infrastructure is how these systems are stewarded; it is the invisible architecture that keeps technology aligned with purpose. It ensures organizations have the appropriate staff expertise and resources to fulfill that purpose, including being able to interpret outcomes from hard infrastructure. This is critical for differentiating between a technical system problem and a capacity or decision-making problem.
Soft infrastructure includes:
People and roles: Clear authority and accountability for decisions, change management, and maintenance.
Processes and routine: Learning loops, reflection practices, coordination.
Governance: decision pathways, escalation channels, cross-functional alignment, permissions, consent and approvals where necessary, adherence to any legal requirements.
Use and adaptation mechanisms: ways organizations use, manage, and adjust systems as strategy evolves.
Stewardship
The ongoing, resourced care required to update, maintain, and improve tools after launch.
Soft infrastructure is the leadership system that allows technology to live and adapt. Without it, organizations end up with systems no one uses, cannot change, or that no longer match organizational reality.
How the sector built hard infrastructure without soft infrastructure
Over the last decade, funders and partners poured resources into hard infrastructure. However, they did not invest in the leadership to run them, nor did they assess whether their staff were capable of using them appropriately.
Several factors reinforce the imbalance. Hard infrastructure failures are visible, while leadership failures are quiet and distributed across people, teams and projects. On top of that, project-based funding constrains timelines and excludes ongoing maintenance, and compressed build cycles reward launch over durability. Limited digital governance experience leaves boards unsure how to set expectations, and siloed decision-making prevents alignment between program, data, and operations.
The result is predictable and widespread across the sector: we have tools that inspire excitement at launch, but are not used to their full potential, or stop being used entirely. Systems decay because no one is responsible for adapting them and any updates require major overhauls.
Why AI raises the stakes for nonprofits, funders, and companies
Adding AI into the mix accelerates the factors that create an imbalance between hard and soft infrastructure investment. AI compresses decision cycles, increases opacity, and amplifies the consequences of weak governance.
The predictable results will continue, only faster, across nonprofit organizations, funders, corporate impact teams and boards:
Nonprofits risk embedding inequity, automating flawed assumptions, or relying on systems they cannot evaluate or adjust.
Funders invest in tools that cannot be maintained or scaled, creating one-off products rather than durable capability.
Companies deploy tech-for-good solutions into environments that cannot support them, undermining adoption and impact.
Boards face new fiduciary responsibilities related to data risk, AI ethics, capacity, and long-term sustainability, yet few understand the soft infrastructure they must oversee.
While always critical for projects to actually work, soft infrastructure is now a core risk-mitigation layer across all four groups. Without it, digital and AI efforts expose organizations to strategic and operational risk.
What boards, funders, and companies must do now
Boards
Boards must treat soft infrastructure as part of fiduciary duty. As part of their oversight of digital and AI initiatives, they need to require clear roles and decision-making structures for these initiatives, as well as the organizational and staff capacity to operate these systems, and the routines for learning, adaptation, and ongoing alignment with their mission. Boards essentially must become soft infrastructure themselves.
Funders
Funders should treat soft infrastructure as a legitimate and necessary investment category and explicitly fund governance design, decision-making roles, learning routines, and long-term stewardship alongside technical builds. Without this, philanthropy continues to subsidize short-lived systems while claiming innovation.
Companies
Corporate impact teams should evaluate nonprofit partner readiness using a soft infrastructure lens. Product donations, data platforms, and AI tools all require decision pathways, staff capacity, governance, and ongoing coordination. Partnerships succeed when companies help strengthen these conditions and don’t just provide tools.
A soft infrastructure assessment: five questions that predict durability
The five questions below offer a practical way for leaders to assess whether the foundations for responsible, sustainable technology are in place:
How clearly have we defined the people, roles, and decision pathways that guide this initiative as it evolves?
What ongoing routines will help our organization learn from data, staff experience, and context, and adjust accordingly?
How will we maintain, improve, and resource this product or system after launch, not just during the build?
How will leaders ensure ongoing alignment with strategy, staff capacity, and organizational purpose over time?
If strategy or context shifts, how might we adapt without requiring a full rebuild?
Soft Infrastructure is now core strategy
The social sector has never lacked tools. What it has lacked are the leadership systems that allow those tools to produce meaningful results, evolve over time, and stay aligned with mission. In an AI era—when decisions are faster, opacity is greater, and mistakes scale instantly—the costs of weak soft infrastructure are no longer slow or containable.
Strong soft infrastructure is now the foundation of resilience and long-term impact. It is what allows organizations to use technology wisely, adapt as conditions change, and build systems that serve their purpose, rather than drift away from it.
The path forward is clear: organizations that invest in soft infrastructure will learn faster, act with greater confidence, and ultimately amplify their impact. Those that don’t will continue to see digital and AI efforts stall, or worse, amplify risks they never intended to create.
Soft infrastructure has never been optional. But in this moment, it has become the leadership system on which the future of digital and AI work in the social sector depends.