AI With Purpose: As Funders and Nonprofits Adopt AI, What’s Missing?

“If equity is the goal, the nonprofit sector must be at the [AI] table from the start. Without us, equitable AI won’t happen. With us, it’s possible.”

These are words shared by a nonprofit leader who responded to the Center for Effective Philanthropy’s (CEP) nationally representative survey of more than 400 nonprofit leaders earlier this year. The use of artificial intelligence in the nonprofit and philanthropic sectors is not coming — it’s already here.

CEP’s new report, “AI With Purpose: How Foundations and Nonprofits Are Thinking About and Using Artificial Intelligence,” reveals that the majority of nonprofit and foundation leaders are already using AI in their work. This statistic is in line with research conducted by organizations such as Project Evident and Taproot Foundation.

However, despite this — and the fact that more than 90 percent of organizations reported interest in increasing their use of AI — nonprofit and foundation leaders broadly report sharing a similar set of concerns about AI, from the potential for misinformation or inaccurate results, the security and privacy of data, and bias or discrimination in AI algorithms.

Alongside these concerns, nonprofits and foundations alike must also grapple with AI’s potential impacts on the issues and communities they seek to benefit through their work. CEP’s data indicates that it is still uncommon for foundations and nonprofits to be discussing or using AI in a way that promotes equity.

Very few funders told us they provide some type of support to their grantees for AI implementation — and most of these leaders say they never or only occasionally have conversations about ‘equitable AI’ at their organizations, defined as the ethical development, deployment, and use of AI systems to promote fairness, inclusivity, and justice. Similarly, the majority of nonprofits indicate that their organization never engages in conversations about equitable AI.

Despite the relative rarity of equitable AI conversation and implementation in the sector, we do see examples in the field of nonprofit organizations discussing, developing, and using AI tools with equity in mind. In interviews with CEP, leaders in the sector also speak to the importance of “using AI to support marginalized and underrepresented communities” and “making sure AI works as an equalizer rather than telling people not to use it.”  

While our findings suggest there is room for growth when it comes to putting equity at the center of AI conversations, they also reveal that, perhaps first and foremost, there is a gap in understanding and communication between funders and nonprofits when it comes to AI needs. Funders’ understanding of their grantees’ AI-related needs lags — indeed, most nonprofits believe that none of their funders understand their needs and concerns regarding AI.

“Only one funder understands it enough to see how it could be useful. No other funders understand it or have even expressed a willingness to understand it,” one nonprofit leader told us. The vast majority of foundations — 90 percent — also do not offer any kind of support to their grantees for the purpose of adopting AI in their work.

AI is still a new tool to many in the philanthropic and nonprofit sectors, but further developments of AI technology are just over the horizon. Some have speculated these advancements will be a boon to nonprofit organizations as they struggle with ongoing capacity, budgetary, and resourcing constraints.

Ultimately, ensuring that diverse voices and perspectives are included at the decision-making table when it comes to AI development and deployment will be critical for those interested in ensuring AI tools are built to be as equitable and ethical as possible. As one interviewee in CEP’s report aptly shared on the subject: 

Seara Grundhoefer is a senior analyst on the Research team at CEP. Find her on LinkedIn.