Where do foundations stand on AI-generated grant proposals?

Generative AI technologies have made it possible for users to transform raw information—quickly and easily—into a presentable and readable form. Although there is concern about how generative AI affects the integrity of written work and images, there is also hope that responsible adoption can boost the social sector’s impact. Greg Stanley at PEAK Grantmaking has argued that generative AI tools can streamline the grant application process by enabling grantmakers to summarize, categorize, and pre-screen applications and by making it easier for grantseekers to fill out application forms. 

Where do foundations stand on AI-generated grant applications? As part of Candid’s annual survey of foundation giving trends,ⁱ we posed two questions about generative AI: Have you received grant applications created by generative AI? Does your foundation accept, or plan to accept, grant applications with content created by generative AI? For the purposes of this analysis, we’re using “AI-generated,” “created by generative AI,” and “with content created by generative AI” interchangeably. 

57% of grantmakers don’t know whether they’ve received applications created with generative AI

Over half (57%) of survey participants said they didn’t know whether they’ve received grant applications created by generative AI. Most of the comments from respondents who didn’t know indicated they doubted they’ve received AI-generated proposals but there was no way to know for sure. One funder wrote, “Program Officers have received some applications they suspect were created with AI, but no applicants have explicitly stated it.”  

Another 42% of respondents believed they have not received applications created by generative AI. Several indicated their foundations’ application process involves conversations rather than written materials. Only four survey participants, or under 1%, said they’ve received AI-generated applications.    

10% of funders accept grant applications with content created by generative AI

One in 10 respondents indicated they would accept grant applications with content created by generative AI. Comments indicate that many of these respondents believe the broader use of generative AI technologies is inevitable and that, as long as grantees accurately describe their mission and their work, is ultimately of little consequence. Here’s a sample of their comments: 

“We do not have the have the resources or expertise to determine what is AI-generated or not. Additionally, as a community foundation, we know our local nonprofits and do not have a very complicated application. The risk of AI unduly influencing our evaluation process is minimal.” 

“We are agnostic to the tools that applicants use to create grant applications. Whether assisted by AI or not, applicants need to articulate their needs and alignment with our areas of giving in order to receive favorable assessment.” 

Echoing the sentiments of some in the social sector, a few respondents welcomed AI—hoping it would make the application process easier for grantees: 

“We fund a community with a large number of refugees and other non-native English speakers. We are hoping this will help them level the playing field.”  

“We want the grant application process to be as easy as possible. If AI helps accomplish that, so be it.” 

Notably, some of the foundations that responded “yes” to this question also indicated they’re planning to ask grant applicants whether they used generative AI in composing their applications. 

Two-thirds haven’t decided whether to accept AI-generated grant applications 

Another 23% of respondents said they would not accept AI-generated applications. Some of these organizations, however, also indicated their grant applications were by invitation only. Others emphasized that applications are only a part of the grant process and they seek out direct knowledge of grantees’ work before approving grants: 

“We currently don’t have a policy in regards to AI-generated proposals. However, our due diligence requires reviews of financials and site visits where AI impact would be limited or neutralized.”  

“Our associations with our grant partners are very much relation-based.” 

Notably, a full two-thirds of respondents (67%) are still undecided about whether to accept AI-generated applications. Comments from these funders tended to say they have not yet discussed the matter or have not yet established an official policy about AI-generated applications. Even among respondents who plan to accept AI applications in the future, some commented that they have yet to adopt an official policy.  

These survey results suggest that much is yet to be determined about the use of generative AI technologies for grant applications. At the same time, many foundations focused on the benefits of these technologies in making the application process easier for grantees. If these technologies can be adopted responsibly and ethically to ease the burden on grantees, they may help contribute to a more efficient, effective, and equitable social sector. 

Photo credit: Laylabird via Getty Images

ⁱThe 2024 Foundation Giving Forecast Survey was emailed to 2,156 U.S. foundations. A convenience sample of 527 foundations responded, a 24% response rate. Not every foundation responded to every question. The survey focuses on larger foundations, as their giving accounts for a vast proportion of the overall grant dollars awarded.