
Artificial intelligence is creating new opportunities for nonprofits—and new questions for their leaders.
Where can AI actually help? What information should never be entered into an AI tool? How much should we trust AI-generated content? How do we introduce AI without creating unnecessary risk or anxiety for staff? And how do we make sure AI supports our mission rather than distracting from it?
This course will help you answer those questions. You don’t need a technical background—this is a leadership course, not a technology course.
Throughout the course, you’ll use the GGA Responsible AI Adoption Framework:
Understand → Evaluate → Protect → Prepare → Lead → Adopt
Along the way, you’ll complete practical tools that help you start shaping an AI approach for your own organization.
The agenda for this course is as follows. The module build on each other so please complete them in order. Be sure to utilize the 5 downloadable tools and complete the 6 activities. Please note that activities do not need to be turned in.
Introduction - Lectures 1-2
Module 1: Understand - Lectures 3-4
Activity: Identify Where Your Organization Spends its Time
Module 2: Evaluation - Lectures 5-7
Activity: Screen One AI Use Case
Download: MVRR AI Use Case Screening Worksheet
Module 3: Protect - Lectures 8-10
Activity: Question an AI Answer
Download: 5 Questions Before You Trust an AI Answer Worksheet
Module 4: Prepare - Lectures 11-13
Activity: Document AI Boundaries
Download: Responsible AI Traffic Light Worksheet
Module 5: Lead - Lectures 14-16
Activity: Prepare for AI Adoption Conversations
Download: AI Adoption Conversation Planner
Module 6: Adopt - Lectures 17-19
Activity: Build Your AI Pilot Plan
Download: 30-Day Responsible AI Pilot Plan
Conclusion - Lecture 20
As you move through the course, keep a real nonprofit organization in mind—ideally your own—and hold on to three questions:
Where could AI create meaningful value?
Where could AI create unnecessary risk?
What would responsible AI adoption look like for our organization?
AI is moving quickly, and nonprofit leaders hear everything from “AI will transform your organization” to “AI is too risky to use.”
Your job isn’t to become an AI expert. It’s to understand AI well enough to make good decisions about it. In this module, focus on separating capability from hype.
Generative AI can help people create, organize, summarize, analyze, and brainstorm. But it can also produce incorrect information, miss important context, and deliver answers that sound far more authoritative than they are.
That’s why human judgment remains essential.
Identify where your organization spends its time before deciding where AI belongs.
Step 1: List the 5-10 tasks that take the most staff time each week or month.
Step 2: For each task, note how often it happens, how long it takes, and who does it.
Step 3: Mark the tasks that are repetitive, draft-heavy, or difficult to scale.
Step 4: Flag any that involve confidential, donor, client, or employee information.
Step 5: Circle 2-3 tasks where a better first draft or summary would free time for mission work.
Don't decide yet whether AI should take on any of these. Keep your list - you'll use it in Module 2 to screen one AI use case.
Key takeaway: The goal isn't to find everything AI can do. It's to understand where AI might help your organization accomplish its mission more effectively.
A new AI tool can be impressive without being particularly useful to your organization.
So instead of asking, “How can we use AI?” start with “What problem are we trying to solve?” Then decide whether AI belongs in the solution.
Use the MVRR screen to evaluate any potential AI opportunity:
Put the MVRR screen into practice using the MVRR AI Use Case Screening Worksheet (in this lecture's Resources).
Step 1: Revisit the 5-10 tasks you identified in Lecture 4.
Step 2: Choose one task and complete the MVRR AI Use Case Screening Worksheet: define the use case, score it on Mission, Value, Risk, and Readiness, and record a leadership decision.
Step 3: Answer the reflection questions:
- Which section felt the most natural to complete, and which made you pause?
- Did filling out this worksheet change your initial opinion of the AI idea?
- Would completing it be different if others had been involved?
Key takeaway: The best pilot candidates combine high mission and value with manageable risk and sufficient readiness.
Nonprofits run on trust. Clients trust you with personal information. Donors trust you with resources. Employees trust you with their data. Communities trust you to act responsibly. That trust should stay at the center of every AI decision.
Before relying on AI-generated information—or letting staff enter organizational information into an AI tool—consider six areas:
Keep the Five Questions Before Trusting an AI Answer reference (in this lecture’s Resources) on hand when reviewing AI-generated work—and consider sharing it with staff who are already experimenting with AI.
Key takeaway: AI can assist with a task, but it shouldn’t remove accountability. A person should remain responsible for consequential decisions.
Practice the Five Questions on a real AI answer using the Five Questions Before You Trust an AI Answer Worksheet.
Step 1: Open an AI tool and enter this prompt:
"Act as a nonprofit management consultant. Our nonprofit organization wants to increase volunteer participation by 25% over the next six months. We have limited staff capacity and a small budget. Recommend five strategies to achieve this goal. For each strategy, explain why it should work, identify expected benefits, and suggest how we could measure success. Include any relevant statistics or research that support your recommendations."
Step 2: Use the AI's response to complete the worksheet: Evidence, Assumptions, Context, Error and Bias, Human Check, and the Verification Record.
Step 3: Answer the reflection questions:
- How confident were you in the AI answer before completing the worksheet?
- What surprised you most when you began examining the AI's recommendations more critically?
- How might this experience change the way you or your team evaluate AI-generated information in the future?
Key takeaway: Fluency is not evidence. High-impact uses require stronger human review, not just better prompts.
There’s no universal Green–Yellow–Red list. Your boundaries should reflect your mission, stakeholders, information, regulatory environment, technology, and risk tolerance.
Green (generally OK): public or non-sensitive information, low consequence of error, human review before use, approved tools.
Yellow (check first): internal information, external-facing content, higher reputational impact, unclear ownership or sourcing.
Red (do not proceed): sensitive personal or client data, automated high-impact decisions, unapproved tools, no meaningful human review.
Download the Responsible AI Traffic Light List Worksheet from this lecture's Resources. You'll complete it in the next lecture, Activity: Document AI Boundaries.
Key takeaway: Good AI governance doesn’t just tell people what they can’t do. It helps them understand what they can do responsibly.
One of the biggest risks isn’t that employees use AI—it’s that they use it without knowing what is and isn’t appropriate.
You may not need a full AI governance program right away, but your people do need guidance. A simple place to start is a Green–Yellow–Red framework:
Put your Traffic Light List into practice using the Responsible AI Traffic Light List Worksheet (see Lecture 12 Resources).
Step 1: Document examples/rules for your organization in the Green, Yellow, and Red categories (aim for at least three of each).
Step 2: Customize your boundaries by documenting approved AI tools, required approvals, disclosure expectations, and who owns questions and updates.
Step 3: Answer the reflection questions:
- What surprised you most as you worked through the worksheet?
- Did completing the worksheet change your perspective on the potential benefits or risks of AI in your organization? How?
- What is the most important gap or opportunity you identified through this exercise?
Key takeaway: Good AI governance doesn't just tell people what they can't do. It helps them understand what they can do responsibly.
Introducing AI isn’t just a technology initiative—it changes how people work.
Some employees will be enthusiastic. Others will be skeptical. Some may worry AI will replace their jobs. Others may already be using it without telling anyone. Leaders need to make room for all of these reactions.
Instead of opening with “We’re implementing AI,” open with the problem: “Here’s a challenge we’re trying to solve.” Then explain why AI is being considered as one possible solution. This keeps the conversation focused on organizational needs, not technology for its own sake.
Prepare for a real AI adoption conversation using the AI Adoption Conversation Planner (use it before a staff meeting, leadership meeting, board discussion, or AI pilot rollout).
Step 1: Complete the five sections of the planner:
- Why: What problem, capacity gap, or mission need could AI address?
- Where: Where could AI help? Where is it inappropriate?
- How: What tools, data rules, review steps, disclosure, and accountability will guide use?
- People: What concerns may staff have about roles, workload, fairness, or learning?
- Learning: How will we share learning and respond when things go wrong?
Step 2: Answer the reflection questions:
- Are you feeling more confident about leading an AI adoption conversation now that you've completed this planner?
- Which sections of the planner were easiest and most difficult to answer?
- How effectively does your planned conversation address employee concerns, encourage honest questions, and create an environment where people feel safe discussing mistakes?
Key takeaway: People need both freedom to raise concerns and clear rules for using AI.
Employees may wonder:
You don’t need every answer on day one. You do need an environment where people feel safe asking.
Before introducing an AI pilot or initiative, complete the AI Adoption Conversation Planner (in this lecture’s Resources). It walks you through:
Key takeaway: Successful AI adoption depends as much on leadership, communication, trust, and change management as it does on technology.
You don’t need to transform your organization overnight. Choose one useful, manageable AI application and test it. A focused pilot lets your organization learn while limiting risk.
Your pilot should answer seven questions:
A successful pilot isn’t one where everyone finds the technology impressive. It’s one that produces meaningful results. Consider measuring:
At the end of the pilot, choose one of four paths:
Stopping a pilot isn’t failure. Learning that an application doesn’t fit your organization can save a much larger investment of time, money, and trust.
Now complete your 30-Day Responsible AI Pilot Plan and Measurement Sheet (in this lecture’s Resources). It brings together everything you’ve built in this course.
Key takeaway: Start small. Learn deliberately. Measure results. Expand only when the evidence supports it.
Turn one promising AI use case into a controlled 30-day learning experiment using the 30-Day Responsible AI Pilot Plan.
Step 1: Complete the pilot plan:
- Define the pilot: workflow, problem, mission value, owner, users, and dates.
- Measure success: compare baseline and results across efficiency, quality, risk, people, and mission.
- Set guardrails: tool, data rules, human review, approvals, disclosure, and escalation.
- Check each week: Week 1 define, Week 2 test, Week 3 learn, Week 4 review.
- Make a decision: use evidence to stop, revise and retest, continue, or scale, and name the next action, owner, and date.
Step 2: Answer the reflection questions:
- How confident do you feel about planning an AI pilot after completing this worksheet?
- Did completing the worksheet reinforce your confidence in the AI use case you selected? Why or why not?
- How effectively does your plan involve the people who will use AI?
Key takeaway: Measure value, not excitement. Stopping a pilot isn't failure.
You now have a practical framework for approaching AI as a nonprofit leader:
Artificial intelligence is creating new opportunities for nonprofits—and new questions for the leaders responsible for them.
Where can AI genuinely help? What information should never go into an AI tool? How much should you trust AI-generated content? How do you introduce AI without creating unnecessary risk or anxiety for your team? And how do you keep AI in service of your mission rather than a distraction from it?
This course gives you a practical way to answer those questions. It is a leadership course, not a technology course—no technical background required.
You’ll work through the GGA Responsible AI Adoption Framework, six steps that take you from first questions to a tested pilot:
Understand – separate AI’s real capabilities from the hype
Evaluate – screen potential use cases with the MVRR screen: Mission, Value, Risk, Review
Protect – safeguard confidentiality, accuracy, fairness, and the trust your organization has earned
Prepare – give staff clear Green–Yellow–Red boundaries for AI use
Lead – manage AI adoption as the people change it is
Adopt – run a focused 30-day pilot and measure whether it works
Every module comes with a downloadable worksheet. By the end of the course, you’ll have assembled the foundation of a 30-Day Responsible AI Pilot Plan for your own organization.
Your instructor, Natalya Bah, is the founder of The Growth Group Academy, LLC and advises executive leaders on AI adoption, organizational transformation, and leadership agility. She holds the PMP and PMI-CPMAI credentials from the Project Management Institute and is a part-time instructor at Graduate School USA and Montgomery College.
You don’t need to become an AI expert. You need to understand AI well enough to make good decisions about it. This course will help you do exactly that.