The association, the foundation, the federation

Essay #006 · 2026 · Reading time 6 min · Field: AI Strategy · Market: Sport / NGO

A foundation director I have known for several years recently asked me a question. She said: everything written about AI is about productivity. Faster work, lower costs, higher margins. What about us? About organizations whose purpose is not margin, but impact. I will call them values-based organizations here.

The question has two levels. The first: are there sensible agent uses for associations, foundations, and federations too? The second, deeper one: does AI change what a values-based organization is at its core? This essay answers both.

Why “this is not for us” falls short

In the conversations I have with association, foundation, and federation leaderships, the first thing I often hear is: “This is not for us.” The reasons: tight resources, data protection, skepticism toward technology, the worry that AI displaces human attention. Each of these concerns has a true core. The conclusion is still wrong.

It assumes that using AI always means what it means in companies. That is not the case. The technology is the same, the task is not. In companies, the core question is: how do we create more value with less effort? In values-based organizations, it is: how do we create more impact with the same effort? That sounds like wordplay. But it leads to different uses, different standards, and different decisions.

Four fields where agents can help your organization

I see four fields where agentic work becomes relevant for values-based organizations. Not every field matters equally for every organization. But I expect this: every organization with more than twenty full-time staff will face all four in the coming years.

Field 1: Reach. Getting to the right people. The organizations I talk to rarely fail for lack of topics. They fail for lack of outreach. A support association for rare diseases often reaches the people affected only by chance. Agents can find target groups, address them individually, and strike the right tone. Such work used to require large budgets. A small foundation can build a reach that used to be reserved for large organizations.

Field 2: Impact measurement. Proving what you achieve. The question “What impact do we actually have?” is hard to answer because impact is diffuse, slow, and qualitative. Agents can collect and analyze data that there was no capacity for until now: participant trajectories, interview analyses, media coverage. That does not replace the qualitative work. It complements it and makes you more accountable to funders and members.

Field 3: Administration. Freeing up time for the substantive work. In your organization too, administration ties up people who would rather work on substance: grant applications, reports, communication with funders, documentation. Agents can take over part of that work. The questions are the same as in companies. Only the goal is different: not more margin, but more time for your organization’s purpose.

Field 4: Ethics. Becoming a voice in the debate. This is where values-based organizations can be ahead of business. When agents make decisions that affect people, ethical questions arise. In the companies I know, these questions are usually checked off as a compliance item. Values-based organizations can work on them seriously, because that is their home ground. Those who speak up early help shape a debate that others will otherwise shape.

Does AI change what your organization is?

Beneath the four fields lies the question the foundation director was really getting at. Values-based organizations emerged as a counterweight to economic logic. They stand for things without a market price: dignity, participation, memory, justice. Agent systems, by contrast, are tools of exactly that economic logic, built for efficiency and scaling. Does that fit together?

My answer: yes — but only if the leadership takes responsibility itself. Those who introduce agents without making their own values the standard will water those values down. Those who make their values the standard for every deployment achieve the opposite: an organization that knows more clearly what it is.

You cannot delegate this. The first five essays in this archive describe the same thing for companies: the decision-maker has to work through the topic personally. A director who leaves it to her IT department is not leading her organization through this change — she is watching. A federation that relies on an external advisor will not notice which of its principles are currently up for grabs.

The question “Does AI belong to us?” is therefore misleading. The right question is: how does our organization change when we use AI, and do we want that change? You answer that question, not your IT and not an advisor.

A concrete first step: put the topic on the agenda of your next board or committee meeting. Not as a technology item, but as a values question: which of our tasks may agents take over, and which never?

What arocon offers values-based organizations

arocon works with values-based organizations on similar terms as with companies. There are two differences.

First: each year I take on two engagements at reduced rates or pro bono, selected by relevance and fit. That is not charity, but part of my own values position.

Second: the questions we discuss are usually different: less margin, more impact. The posture is the same: the leadership should be able to decide for itself what happens in its organization, and why.

If you want to make that decision rather than have it happen to you, this is the moment to start.

— Axel Roth