Scaling AI in not-for-profits must preserve trust

Why experimentation is not enough

Urgency alone does not create value. Moving from potential to tangible results requires discipline. Greg Haberer, Grant Thornton Associate Partner in Risk Advisory, cautioned that simply deploying AI tools does not guarantee better outcomes.

“All we may have done is spend more money to get the same work done. So, what is the plan?” Haberer said. “Any investment in AI needs to be paired with a clear plan for what you’re executing from a business operations standpoint.” Nauman Afzal, Associate Partner, Technology Advisory, adds that AI’s impact depends as much on integration and execution as on the tools themselves.

For many NFPs, the challenge is shifting from gaining access to AI tools to organizing them into a coordinated strategy that delivers measurable value. Without defined ownership, prioritized use cases and clear decision rights, AI activity expands but enterprise value does not. Early gains across fundraising, operations and service delivery remain uncoordinated rather than connected to enterprise priorities. This results in fragmented data, duplicate investments and inconsistent governance that make it difficult to scale AI beyond individual use cases.

NFPs do not need more experimentation. They need the discipline to organize, govern, and connect AI use to measurable mission and operating outcomes.

Where AI is already delivering value today

Early adoption patterns are becoming clearer. Not-for-profits are applying AI first in areas where benefits can be demonstrated quickly, including fundraising, stakeholder engagement, operational performance and mission delivery. Historically, limited resources constrained organizations’ ability to personalize outreach at scale, but AI is helping to make that more possible.

“Now NFPs can use AI to create finely tuned, personalized marketing outreach and reach long-tail audiences.” said Crystal Wang, Grant Thornton Director, Advisory AI-ML Innovation. “Staff can craft stories that resonate emotionally, which is critical when people decide to donate.”

These capabilities enable more targeted engagement, strengthen donor relationships and improve the efficiency of outreach efforts, helping organizations generate greater returns.

Organizations are also applying AI across human resources, finance and administrative functions to streamline routine work, improve productivity and support faster decision-making.

“There are already many use cases in the market, and NFPs can harvest what has already been done in other industries, particularly in the back office,” Wang added. Even more significantly, AI is beginning to extend into mission delivery itself. Applications such as fraud detection, personalized learning experiences and constituent support tools are enhancing service delivery, expanding reach and improving responsiveness.

Governance as a foundation for scale

As these use cases begin to move closer to beneficiaries and sensitive data, they introduce new complexities. Trust, long a defining feature of the NFP industry, sits at the center of this challenge.

As AI becomes more embedded in select workflows and functions, trust can no longer remain a value statement. It must be translated into a set of decisions leaders make and stand behind:

• What data can be used—and under what conditions
• Who is accountable for AI-driven decisions
• How outputs are validated and explained
• How AI use aligns with mission commitments and stakeholder expectations

Without this clarity, organizations risk scaling AI faster than they can explain, govern or defend it. NFP organizations cannot integrate AI without guardrails, Haberer said. Leaders must understand who has access to data, how that data is used within AI-driven processes and what controls are in place to manage risks.

Governance is the foundation that determines whether AI remains confined to isolated pilots or can evolve into a trusted enterprise capability integrated into core operations over time. According to Wang, effective governance operates across multiple layers. Data governance establishes what information can be used and under what conditions. Solution governance addresses how AI systems are designed, deployed and maintained. Process governance helps ensure outputs are reviewed, validated and monitored, particularly in high-stakes scenarios.

Beyond risk management, Afzal positions governance as a visible and strategic capability. He notes that NFPs can strengthen trust by proactively communicating how AI is governed: through ethics committees, usage guidelines and public-facing transparency around data practices. Making governance explicit not only mitigates risk but reinforces credibility with donors and stakeholders, enabling organizations to scale AI with greater confidence.

“It’s not a commercial transaction,” Afzal said. “These are individuals donating their hard-earned money. NFPs must be fully transparent and protect donor information because those relationships are built on trust.”

Together, these layers transform trust from an organizational value into an operational capability. Organizations that build them deliberately can scale AI across the enterprise. Those that do not often find promising pilots stalled because the governance required to scale it was never established.

Embedding governance into enterprise operations

Even when AI demonstrates high levels of accuracy, human oversight remains essential for many use cases. Organizations must balance automation with accountability, ensuring that critical decisions remain aligned with mission objectives, stakeholder expectations and organizational values. When this happens, leaders can make informed decisions about where AI can create value, what risks are acceptable and how accountability should be distributed across the enterprise.

These perspectives point to a broader shift in how NFPs approach governance. Historically, governance has been focused on compliance and oversight. Increasingly, it is becoming embedded within day-to-day operations, which Haberer said encourages organizations to assess AI applications based on levels of risk and align controls accordingly.

This shift is critical because governance must be a foundation for scale. Without it, AI initiatives remain fragmented, with function-specific gains across fundraising, operations or service delivery that fail to translate into meaningful enterprise-wide impact.

Today, many AI efforts are disconnected across the organization. Scaling AI requires progression and at each stage, leadership teams face decisions that are often deferred:

• Understand the landscape ꟷ Identify where AI is already being used across fundraising, operations and mission delivery.
• Prioritize by risk and value ꟷ Focus on use cases with the greatest impact and the highest governance requirements.
• Establish accountability ꟷ Define ownership for oversight, performance, compliance and business outcomes.
• Measure what matters ꟷ Track operational efficiency, mission impact, stakeholder trust and financial results.

The process begins with visibility. Many organizations cannot effectively govern AI because they lack a clear understanding of where it is already being used. “You have to start with your use cases,” Haberer said. “You need to ask, ‘What is our use-case inventory, and where do we think AI can actually be applied?’”

From there, success depends not only on identifying opportunities, but also on tracking performance and adapting strategies over time. Wang added that these efforts must be supported by readiness across data, talent and technology, creating the conditions that support more consistent and sustainable value creation over time.

Achieving this level of maturity requires cross-functional leadership. Sustainable adoption depends on alignment across business, technology, risk, and operations. This alignment ensures investments are prioritized, accountability is clear, and initiatives are guided toward mission-driven outcomes.

The governance model below illustrates how governance and oversight requirements evolve as AI adoption expands and related risks increase.

Show image description -->

Turning experimentation into enterprise impact

For NFP leaders, the path begins with understanding where AI is already in use and identifying high-impact opportunities. It extends to strengthening data foundations, implementing clear governance frameworks and building processes for continuous review. Just as importantly, it requires aligning AI initiatives with mission priorities and stakeholder expectations.

Afzal notes that moving beyond early experimentation depends on modernizing underlying systems and aligning investments with long-term strategy. Many NFPs remain constrained by legacy platforms, and organizations that prioritize modernization and enterprise alignment will be better positioned to move from isolated use cases to sustained impact.

AI introduces powerful new capabilities, but it does not change the fundamental purpose of NFPs. As Morrone noted, core priorities ꟷ engaging stakeholders, delivering mission outcomes and maintaining trust — must remain unchanged to preserve integrity and long-term value. AI is simply a new means of advancing those enduring objectives.

The NFP sector’s AI journey is still emerging. Those that begin building this operating discipline now will be best positioned to turn early experimentation into sustained mission impact.