Executive summary
AI has seen widespread initial adoption in the nonprofit sector, but many organizations struggle to avoid common pitfalls and lack the resources to optimize their investments. Through intentional adoption and sound governance, not-for-profits can ensure AI serves their mission and the communities they support.
Not-for-profits have a distinctive relationship with generative AI. Grant Thornton’s Head of Non-Profit and Higher Education Dennis Morrone explained that “not-for-profits understand the potential of AI — or, at least, they’ve been sold on the potential. But they haven’t yet been able to clearly see the ROI or assess the efficacy of their initial efforts.”
Many not-for-profits have begun using AI, though, without a clear long-term strategy or understanding of how it fits into their broader mission. Not surprisingly, over 80% have adopted AI in some way. They sense that enthusiastic adoption could give them a competitive advantage in everything from donor engagement to operational efficiency. Conversely, sluggish adoption could leave them behind their peer organizations.
But ill-advised implementations also risk compromising many of the qualities that set not-for-profits apart: equity, trustworthiness, accountability and a respect for donors’ and beneficiaries’ privacy. For example, fundraising algorithms trained on historical donor data can disproportionately target wealthy, older, less diverse donors. This focus, even if well-intended, can be a detriment because AI systems trained to focus on demographic “safe bets” can unintentionally signal who “belongs” in a donor base and who doesn’t. This can thwart AI’s potential to build broader, more diverse donor coalitions that, ultimately, are stronger and longer-lasting.
Only about 10% of not-for-profits have policies regarding AI’s use and, unlike their for-profit counterparts, many lack the resources and infrastructure to quickly implement such policies. Fortunately, there are steps even the smallest organizations can take to reduce the risks while also optimizing the opportunities.
Start with the data
Because they must do more with limited resources, not-for-profits’ data sources may not be as consistent, rich, or complete as they would want them to be. AI may amplify any biases already inherent in the data. Grant Thornton Technology Modernization Services Partner Zac Taylor said that “inequities, biases, and AI ‘hallucinations’ are often correlated to the maturity of the underlying data.”
Taylor suggests organizations take a hard look at their existing data and ask these questions:
- What are the things that we know to be true?
- What are the things we don’t know about it?
- What purpose was the data collected for?
- What is the likely or stated understanding of beneficiaries or donors about the use of their information?
- Where is the data stored and who has access to it?
- Are we making assumptions that might be creating bias?
Moving forward, businesses should use this assessment to create a process that can lead to cleaner data with less chance for bias or misuse.
Find the best use
For organizations whose staff is already stretched thin, the best use of AI may not be donor engagement. Using generative AI to proactively suggest how much or how often donors should contribute might be better left to the judgment of experienced team members. Likewise, organizations might want to think twice before creating potentially frustrating or harmful beneficiary interactions, such as a call tree or a consultative chatbot.
Especially in an organization with limited staffing, the initial best use may be taking care of routine but critical tasks such as creating meeting minutes, aggregating or summarizing material, providing or gathering generic information, and creating routine communications such as emails.
This delegation would allow staff to spend more time engaging with donors and constituents. As a general rule, the closer a task is to the essential work of the organization, the more important it is for a human to be involved. And, as a general philosophy when starting out, it might be smarter to use AI to empower your staff rather than using valuable staff time to enhance your AI.
Don't wait for regulation
However they deploy AI, not-for-profits should establish policies, procedures and governance before doing so. While it’s tempting to wait for governmental and industry organizations to provide guidance, they appear reluctant to do so for a technology that is quickly evolving and not fully understood.
The approach to governance, including the designated lead, will vary by organization. In larger organizations, this may be a Chief Privacy or Chief Data Officer. In smaller ones, the CFO may champion AI governance initiatives. The content of the guidance will depend on factors such as the organization’s mission; its size, scope, and complexity; its donors and constituents; its workflows and touchpoints; and the nature of its broader regulatory and reputational risks.
At the very least, these guidelines should require clear disclosures whenever AI is used to generate content or make recommendations. They should include some way to iterate guidelines as the organizations learn what works and what doesn’t. Finally, such guidelines should also promote training to reduce the likelihood of unfortunate incidents that often inform subsequent regulations.
Robust prompts and skeptical use
Rudimentary training could focus on two pivotal points in the workflow. The first is writing appropriate prompts to direct the AI. Since these are the starting point for the generative activities, their quality disproportionately shapes what is produced. As Taylor emphasized, “The more context we provide, the more parameters we provide, the richer the data it can access, the more we tell AI to play a certain role or persona, the more sophisticated and trustworthy its answers become.”
No matter how good the output becomes, it’s important that it be actively reviewed. Ultimately, such review is the only way to proactively manage risk and reduce the chances of the kind of mishap that gets organizations unwanted news coverage.
Staying out of the news is a low bar. If they implement AI wisely, in ways that maximize their human resources, not-for-profits can advance their missions as never before.
Key takeaways
- While over 80% of not-for-profits have adopted generative AI, only about 10% have internal governance policies over its use and privacy.
- Not-for-profits have distinct challenges, including constrained resources. Equity, trustworthiness, accountability, and a respect for donors’ and constituents’ privacy are central to their success.
- While AI can amplify inequities in data, the effect can be minimized by a rigorous examination of the underlying donor information.
- The best use of AI may be in eliminating mundane tasks and freeing up staff to engage with donors and constituents.
- Organizations should proactively address governance and training issues related to AI practices.
Contacts:
Partner, Technology Modernization Services
Grant Thornton Advisors LLC
Zac Taylor is a Principal within Grant Thornton’s Organizational and Operational Transformation practice.
Dallas, Texas
Industries
- Asset Management
- Banking
- Life Sciences
- Technology, Media & Telecommunications
- Manufacturing, Transportation & Distribution
- Construction & Real Estate
Service Experience
- Advisory Services
- Operations and Performance
Head of Not-for-Profit & Higher Education Industry
Partner, Audit Services, Grant Thornton LLP
Content disclaimer
This Grant Thornton Advisors LLC content provides information and comments on current issues and developments. It is not a comprehensive analysis of the subject matter covered. It is not, and should not be construed as, accounting, legal, tax, or professional advice provided by Grant Thornton Advisors LLC. All relevant facts and circumstances, including the pertinent authoritative literature, need to be considered to arrive at conclusions that comply with matters addressed in this content.
Grant Thornton Advisors LLC and its subsidiary entities are not licensed CPA firms.
For additional information on topics covered in this content, contact a Grant Thornton Advisors LLC professional.
Trending topics
No Results Found. Please search again using different keywords and/or filters.
Share with your network
Share