RSMUS.comNonprofits today are being asked to do more with limited resources while navigating funding uncertainty, workforce challenges, and increasing expectations from boards, donors, and constituents. At the same time, AI conversations are accelerating. Many organizations feel pressure to act, even if the necessary strategy, governance, or data foundation is not yet in place. That’s why avoiding common pitfalls early can help reduce risk, increase adoption, and generate meaningful value faster.
So far this year, the RSM team has hosted “Building an AI-Ready Nonprofit: From Strategy to Action” events in Charlotte, Philadelphia, New York City, Boston, Toronto, Chicago, and Washington, D.C. During these conversations – and in many follow-up discussions with nonprofit executives, technology leaders, fundraising teams, and program leaders – we have heard a common theme: Organizations recognize AI’s potential, but many are struggling to determine where it fits, how to use it responsibly, and how to move from experimentation to meaningful results.
The good news is that most organizations are facing similar challenges. AI readiness, governance, efficiency, and measurable outcomes continue to be some of the most discussed topics among nonprofit organizations today. Based on the questions, concerns, and lessons we’ve learned from nonprofits across North America, here are five of the most common AI mistakes we see, and practical ways to avoid them.
Mistake 1: Building custom without understanding the tools and agents that you already own
Most modern applications have built-in AI capabilities. Before deciding to build a custom solution, audit your systems and identify the tools you are already paying for. Some applications have agent marketplaces, such as Salesforce’s AgentExchange, where you can even install pre-built agents tailored to your use cases.
What to do instead:
- Inventory your existing AI-enabled tools.
- Identify available copilots, assistants, and agents already included in your software investments.
- Test low-risk use cases before investing in custom development.
Mistake 2: Leading with productivity and stalling on high-impact use cases
While AI-enabled productivity is an easy first step, it often doesn’t radically transform your work. High-impact use cases are often those that change the way your teams work. These use cases streamline your processes and free up your time to deliver more impact.
Struggling to identify or prioritize these use cases? Start by looking at your strategic plan and the blockers you face. Use AI as a tactic to support these strategic goals. Make sure you establish KPIs early to help you evaluate the effectiveness.
What to do instead:
- Start with strategic priorities.
- Identify manual, repetitive processes.
- Define success metrics before implementation.
Mistake 3: Starting complex before the groundwork is ready
Implementing AI is different than traditional technology implementations. It is highly iterative and operates on a much faster pace.
Start small, focusing on one use case for one team, and focus on building skill and trust as you evolve. Early wins help build organizational confidence, create internal champions, and establish the foundation for broader AI adoption.
What to do instead:
- Start with one team.
- Establish a pilot.
- Measure outcomes and iterate.
Mistake 4: Overlooking governance and change management
AI is a new way of working. With it comes heightened need for governance and change management.
As nonprofits, we need to be sure that we are good stewards of our constituents’ data. We also need to be sure that our use of AI is in support of our mission. Governance ensures that AI is used responsibly and thoughtfully.
Change management ensures that your teams trust your AI tools and understand how these tools can help your teams focus on the mission-critical work.
AI governance is not about slowing innovation. It’s about creating the guardrails that allow your organization to scale AI safely and responsibly.
Key governance questions include:
- What data can AI access?
- How will AI-generated content be reviewed?
- Who is accountable for AI outputs?
- What policies guide AI usage across the organization?
Mistake 5: Viewing AI as an IT responsibility
AI is changing the way everyone works. As a result, it is critical that every team has ownership and buy-in for AI adoption. Siloing AI within your IT department will stymie its adoption and potential impact.
AI success requires participation from leadership, program teams, development teams, marketing and communications, finance, operations, and IT. The most successful organizations treat AI as an organizational capability, not a technology project.
What to do instead:
- Create a cross-functional AI working group.
- Include executive sponsorship.
- Engage end users early in the process.
- Establish shared accountability for adoption and outcomes.
Building an AI-ready nonprofit starts with a practical plan
Artificial intelligence isn’t a single project or technology purchase. It is a new way of working.
The organizations seeing the greatest success are not necessarily those moving the fastest. They are the ones taking a thoughtful approach—aligning AI initiatives to mission outcomes, establishing governance, focusing on measurable value, and building adoption one use case at a time.
If you’re evaluating AI for fundraising, program delivery, constituent engagement, case management, operations, or reporting, remember that the goal isn’t simply to adopt AI. The goal is to create more capacity for your people to focus on mission-critical work.
Ready to assess your organization’s AI readiness?
Whether you’re exploring your first AI use case or evaluating how to scale existing efforts, the RSM nonprofit team can help.
- Connect with our nonprofit and Salesforce specialists to discuss your organization’s goals and challenges.
- Request a complimentary Health Check for Salesforce to better understand your current environment and opportunities for optimization.
- Explore additional nonprofit insights, webinars, and upcoming events from RSM. Explore more nonprofit tech blogs here.
Contact our team today to schedule a conversation and learn how your organization can build a practical, responsible, and mission-aligned approach to AI.
Lou Bialon-Crane
Lou is a 16x certified Salesforce architect who provides Salesforce consulting services to public and nonprofit organizations. He has a background in nonprofit operations.
