Many senior leaders in program, evaluation, grants management would prefer not to weigh in on tech, IT, infrastructure discussions. They may talk about technology as if it were a tool you install and move on from. Pick a system. Flip the switch. Train the staff. Problem solved.
Tech, IT, and infrastructure are not set-it-and-forget-it projects. Technology is a critical part of the organizational house we all must navigate and live in.
A house requires design choices, shared norms, ongoing maintenance, and clear responsibility. It shapes behavior and encodes values. When no one agrees on how the house works, mess ensues quickly and expensively.
Anyone who has ever lived with a broken toilet understands this. It was a very real issue last week for one of my collaborators. The fix required a day out of office and partnership with an expert. He had to allocate time, financial resources, and personal capacity to address the problem. But once repaired, the system worked again.
In your organizational house, the fixture itself may be perfectly functional, but if no one agrees on how it is used, cleaned, repaired, or reported when something goes wrong, the problem escalates fast. Technology works the same way. Installing a system without governance does not create efficiency. It creates risk.
Our IT houses demand more of us than procurement decisions.
We Are Already Living in the House
In the social sector, we have experienced digital transformation many times over. The current disruptive moment is being enabled by AI. We cannot plaster over the cracks. With AI-adoption, we have to go to our bedrock of data and culture. We are already living inside systems heavy with accountability and power.
In our Foundation Review article on AI, trust, and grant reporting, my coauthors and I argued that reporting systems failed not because they were inefficient, but because they were designed as mechanisms of control rather than learning. Technology amplified that failure because infrastructure always reflects the values of its builders.
I often use metaphors to describe the nuance AI adoption requires. AI as a Mirror reflects back our culture, our habits, our bias. AI as a Bicycle describes a world where AI-enabled processes will entrench status quo practices unless systems are intentionally designed for equitable and expansive use.
As with home ownership, major investments are often delayed. Rewiring or replumbing a house is expensive and disruptive. Implementing a new grants management system can feel the same. AI squarely falls into this category of major overhaul.
And yet, the warning signs may already be present. Data leaks. Shadow systems. Quiet workaround that slowly push an organization past the point of decision. The questions becomes unavoidable: are we willing to take responsibility for the house we are already occupying?
Who Notices the Cracks First
In most foundations and nonprofits, the people who notice problems first are not senior leaders or boards. They are grants managers.
Grants managers live in the plumbing. They see where systems leak, where workflows clog, where staff and grantees are forced into workarounds. They are often the first to recognize when a tool is creating burden rather than clarity, or when a process quietly undermines stated values.
Grants managers hold enormous power and responsibility precisely because they sit at the intersection of systems, people, and practice. They feel the strain before it shows up in dashboards or board memos.
Senior leaders, by contrast, hold governance and strategy. They approve tools, set risk tolerance, and define what accountability looks like. When leadership abdicates this role or treats technology as a back-office issue, the house deteriorates.
And all of us participate in how values are practiced daily through process. Every shortcut, workaround, or unexamined habit teaches people how the house actually functions, regardless of what the mission statement claims. Culture is the soil under the house. All work begins and ends with culture.
Building for People, Not for Assumptions
Several years ago, I lived in Australia and learned about a government housing initiative intended to support Aboriginal families. The homes were newly built and technically sound. They met policy requirements. They looked good on paper.
But many families had never used a Western-style stove. Some did not know how it worked. In some cases, people built fires in the middle of the living room, not out of negligence, but out of familiarity and necessity.
The government built houses without building relationships. They designed for an imagined user rather than the people who would actually live there. The result was harm, frustration, and failure that could have been avoided through co-creation and listening.
We make the same mistake with technology all the time.
We adopt tools based on vendor demos, peer pressure, or fear of being left behind. We assume users will adapt. We assume training will solve misalignment. We rarely ask whether these are the systems our organization actually needs, or who they are truly built for.
Governance Is Not Control. It Is Care.
When people hear the word governance, they often think of restrictions, policies, and approvals. But good governance looks much more like maintenance, inclusion, transparency.
It asks:
Who is responsible when something breaks? (Risk management)
How do concerns get surfaced without penalty? (Culture)
Who gets to say no? (Decision-making)
How do we repair harm when systems fail people? (Risk mitigation)
In our work on AI and reporting, we emphasized human-in-the-loop oversight not as a technical safeguard, but as a relational one. Governance that works is governance that stays close to lived experience.
Robin Wall Kimmerer reminds us that relationships cannot be forced. They must be reciprocal. Participation alone is not enough if power and benefit still flow in one direction.
The same is true of technology. Systems that extract time, data, or labor without reciprocity will fail, even if they are labeled participatory or innovative.
Risk-Aware, Not Risk-Blind
Living in a house means accepting some risk. Pipes leak. Electricity systems wear down. Things break. Responsible homeowners do not respond by refusing to cook or use the bathroom. They respond by putting smoke detectors in place, agreeing on norms, and paying attention.
Risk-aware technology governance works the same way. It does not mean avoiding AI or new systems. It means being explicit about tradeoffs, consent, and accountability. Does this system support the people inside it? Does it make work more humane? Does it align with our values?
Choosing Stewardship Over Shortcuts
The most dangerous myth in the social sector right now is that technology failure is a technical problem. It is not. It is a stewardship problem. We do not need faster installations or smarter tools. We need clearer ownership, deeper relationships, and the humility to admit when the house is not working for everyone inside it.
Grants managers and other administrative staff embedded in data and relationship will continue to notice first. Leaders must be willing to listen and govern accordingly. And all of us must practice our values daily through the processes we design, tolerate, and defend.
We have accepted care of the house. And for many of us, it’s under strain ... it may be burning. The polycrisis has disrupted the social sector, but we work in this field of trust and relationship because we refuse to sit on the front lawn and cry alone. We turn to Community. We prioritize empathy, collaboration, and a shared vision of a better future.
Now we have work to do.
