Read a sample: 1. Start With Your Community, Not the Tool
The volunteer coordinator of a neighborhood food pantry has a problem that will sound familiar to many community groups: there is more work than time. Each month, someone must write an announcement, answer routine questions, prepare a volunteer schedule, and summarize the committee’s decisions. One evening, a volunteer suggests using an artificial intelligence tool to handle some of it.
The suggestion is reasonable. So are the questions that follow. Could the tool help draft an announcement without receiving information about pantry guests? Would a meeting summary expose personal stories? Who would check the result before anyone relied on it? And if the group saved an hour but made one neighbor less willing to ask for help, would that be a useful trade?
These are not reasons to dismiss AI. They are reasons to begin with the community rather than a product demonstration. A tool can produce an impressive paragraph in seconds, but it cannot decide what trust means to your group or which responsibilities should remain with people. Those choices belong to the people who understand the relationships at stake.
This chapter will help you name the work you want to improve, identify the people a proposed use could affect, and agree on initial boundaries. You do not need a technical background, a paid consultant, or a finished policy to begin. You do need a conversation that makes room for both practical needs and reasonable concerns.
Groups often encounter AI through a specific feature: a button that summarizes a meeting, a writing assistant in an email program, or a service that promises to answer questions automatically. Starting there can make the decision feel like a choice between adopting the feature and falling behind.
Instead, describe the task without mentioning the tool.
“We need an easier way to prepare our monthly event announcement” is a useful starting point. So is “Our volunteer orientation takes too much time to update when procedures change.” Each statement identifies work that matters and invites more than one possible solution.
A possible response might be an AI-assisted draft. It might also be a reusable template, a shorter approval process, a shared calendar, or a clearer division of responsibilities. Sometimes the best answer is to stop producing a document that nobody uses. A group that begins with the task can compare these options rather than treating AI as the goal.
Try to make the problem specific enough that you can tell whether a change helps. “We want to communicate better” is important but broad. “We want to prepare the first draft of a public event notice with less volunteer time, while keeping the date, access details, and tone accurate” gives the group something it can assess.
It also makes the limits of the proposed use easier to see. Drafting a general notice is not the same task as answering a neighbor’s private request for assistance. Both may involve writing, but they involve different information, different relationships, and different consequences if something goes wrong.
The person typing into an AI tool is not necessarily the person most affected by its output. A committee member may use a tool to draft an announcement, but residents must understand it. A volunteer may propose using a tool to organize applications, but applicants may bear the consequences of an inaccurate description. A meeting host may want a summary, while participants may worry about where their words will go.
Before considering a particular tool, ask whose interests are involved. Depending on the task, the answer might include:
- People receiving services or taking part in activities
- Volunteers and staff who provide information or carry out decisions
- Applicants, donors, speakers, or partner organizations
- People mentioned in records who are not present for the discussion
- Readers who rely on translations, accessible formats, or clear instructions
- Future volunteers who will inherit the process
The list does not need to be exhaustive on the first attempt. Its purpose is to widen the conversation beyond convenience for the operator.
For each group, consider what they have entrusted to your organization. Sometimes it is information: a phone number, a story, an availability schedule, or a concern raised in confidence. Sometimes it is something less tangible, such as an expectation that a person will be treated fairly or that a community event will reflect local voices.
A small library, for example, might want help drafting descriptions for a family reading program. The immediate users are the staff or volunteers who prepare the publicity. The affected community also includes families who need accurate information, local partners named in the notice, and people who depend on the announcement being available in a language they understand. A draft that sounds polished but misstates the age range or invents a registration requirement could make the program harder to access.
The question is not whether every affected person must attend every technology meeting. That would often be impractical. The question is how the group will learn enough about their needs and expectations to make a responsible choice.
Most community organizations already have rules about how they treat people, even if those rules are not collected in one document. A mutual-aid group may protect the privacy of people who ask for help. A neighborhood association may promise that residents can raise concerns without being mocked or dismissed. A volunteer-run arts center may work to represent local artists accurately and seek permission…