AI Readiness For Nonprofits Starts with Boring Questions

By: Stacey Segal, COO

AI conversations can get exciting (or overwhelming) very quickly.

New tools. New possibilities. Faster insights. Smarter workflows. Automated summaries. Recommended next actions. Better segmentation. More personalized engagement.

All of that is interesting.

But AI readiness usually starts somewhere much less glamorous.

It starts with boring questions.

Not because the opportunity is boring, but because the foundation matters.

For nonprofits, AI is useful if it can operate on structured, trusted, governed, and connected data tied to real work. Without that foundation, AI can create more confusion faster than ever.

Where Does the Data Live?

Before asking what AI can do, start with where the data lives.

Is donor data in the CRM? Is marketing engagement data somewhere else? Are event interactions stored in another platform? Are gift details, relationships, preferences, and notes consistently maintained? Are there spreadsheets that still function as unofficial systems of record?

AI cannot responsibly support decisions if the organization does not understand which systems hold which information.

Organizations need a clear view of their data ecosystem and where the trusted sources are, how they relate to each other, and which systems are the source of truth for which types of data.

What Data Can Be Trusted?

AI can summarize, interpret, and recommend. But it does not automatically know whether the underlying information is accurate.

If duplicate records exist, AI may read an incomplete donor history. If codes are inconsistent, AI may misunderstand behavior. If relationships are missing, AI may overlook important connections.

Before expanding AI use, ask:

Which data elements are reliable today? Which data elements are inconsistent? Which fields are required for AI but not maintained well? Is my data structured?

Those answers reveal where AI may be useful now and where more foundational work may be needed.

Who Owns the Rules?

AI readiness is not just a technical question. It is a governance question.

Every organization has rules, whether they are documented or not.

What counts as a donor? What counts as active engagement? How should households be handled? Who can change communication preferences? Which relationships should be maintained? What data should never be used for certain types of outreach?

If those rules are unclear, AI will not magically make them clearer. It may simply amplify inconsistent patterns at scale.

A strong AI strategy needs documented business rules, clear ownership, and agreement between business and technical teams.

What Should AI Be Allowed to Do?

Not every AI use case carries the same risk.

There is a big difference between using AI to summarize internal notes and allowing AI to recommend donor outreach, update records, or trigger communications.

Before implementing AI-enabled workflows, nonprofits should define boundaries.

For example:

Can AI suggest an action, but not execute it? Can AI draft content, but require human review? Can AI summarize a donor record without changing it? Can AI flag potential duplicates, but not merge them automatically? Can AI surface insights only from approved data sources?

These decisions matter because they protect donor trust, staff confidence, and organizational accountability.

Where Are Users Losing Time?

The best AI opportunities are often hiding in everyday frustration.

Look for workflows where staff spend time:

• Summarizing long donor histories

• Preparing meeting briefings

• Cleaning import files

• Reformatting data

• Finding related records

• Defining patterns

These are practical opportunities because they connect AI to real operational value.

The goal is not to use AI because it is new. The goal is to reduce friction in work that already matters.

What Needs to Stay Human?

AI can help teams move faster, but it should not replace judgment.

Nonprofits rely on relationships, trust, context, and understanding of mission. A recommendation may be useful, but someone still needs to decide whether it makes sense. A summary may save time, but someone still needs to know whether it is complete. A suggested next step may be helpful, but it should be evaluated through the lens of donor intent, organizational priorities, and stewardship.

AI works best when it supports human decision-making, not when it quietly replaces it.

The Bottom Line

AI readiness is not about chasing every new tool.

It is about understanding your systems, strengthening your data, clarifying your rules, and identifying where intelligence can responsibly support your work.

The most important questions may not sound exciting:

• Where is the data?

• Can we trust it?

• Who owns it?

• What rules apply?

• What should AI be allowed to do?

• Where would it actually help?

Answer those questions first, and AI becomes much more than an experiment.

It becomes part of a thoughtful, governed, and practical technology strategy.