AI Alone Isn’t Enough: Coordination Is What Really Matters

At Fort Collins Habitat for Humanity, Mara Johnson is excited about the potential of an AI agent that communicates with potential donors. The agent already operates independently in stewarding existing donors, and Johnson, the group’s chief development officer, is revamping the position of a departing donor-relations manager to make it much more data-centric. Once she gets the go-ahead from Habitat’s national office to begin soliciting new donors, the AI her organization calls Val will move to center stage.

“We’ll unleash Val to reach out to those donors and invite actual financial contributions to Habitat in Fort Collins,” Johnson says.

The AI agent is a tool from the fundraising technology firm Blackbaud. The Blackbaud Institute, the firm’s research arm, recently conducted a survey of 1,400 nonprofit employees that found that nonprofits could achieve better revenue gains, retain more donors, and boost staff productivity by committing to a deeper, more systemic use of AI.

The report finds that only 10 percent of nonprofits have committed to deep integration of AI, such as creating an AI usage policy, employing paid and secure AI products, and applying the products consistently across the organization. The vast majority of the rest — some 75 percent of respondents — describe their organization’s use of AI as fragmented and individualized, while 12 percent say they aren’t using AI much at all.

The survey found important benefits to integrating AI. The top tier “AI adaptive” organizations save an average of $621 per employee per week using AI, compared with just $503 for the average organization, the report found. The savings were estimated by asking respondents what an hour of their time was worth and how many hours a week they saved by using AI. Those findings might be optimistic. Another study in the for-profit sector found far smaller cost-saving benefits, with two-thirds of workers saving less than four hours a week through AI use. 

At Fort Collins Habitat, Johnson declined to estimate any dollar-based savings from the organization’s AI usage so far. “I wouldn’t say the proof is there quite yet from a numerical standpoint,” she says.

The top tier “AI adaptive” organizations also report achieving greater revenue growth and productivity, says Kate Averett Anderson, a senior content manager at Blackbaud and the main author of the Blackbaud Institute’s report.

“They have established governance in place, they are consistently applying it across the organization, they have well-documented policies, and they are strategically aligning those policies with the way that their organization moves forward,” Anderson says.

Donors See AI Differently Than Charities

Though the majority of nonprofits surveyed are using AI in some form, the report found large gaps between donor expectations about AI usage and what nonprofits are actually doing. The report, which surveyed more than 1,000 donors in addition to nonprofit professionals, found that 56 percent of donors thought it was important to clearly disclose when AI is used in communications — but only 26 percent of nonprofit employees said their organization was doing so.

Sixty-eight percent of donors said it was important to protect sensitive data used by AI systems, but only 36 percent of nonprofit workers thought their employers were using adequate protections.

Charities that have a written policy and a coordinated strategy on AI are more likely to engage in practices that meet donors’ high expectations, says Anderson. In Fort Collins, for example, the Habitat affiliate discloses Val’s role at the top of every email it sends out, allowing skeptical donors to opt for human interaction instead.

“All of these things kind of snowball on one another,” Anderson says. “If you’re able to put governance in place, you’re able to deliver on these things in a lot more predictable way.”

Larger Groups May Benefit More From AI

Many of these benefits may be more likely to accrue to larger and wealthier organizations. They have all sorts of advantages over grassroots charities — including more staff members and perhaps a full-time tech person, says Karen Boyd, an economist at the Policy and Innovation Center who focuses on how charities can use AI in a way that matches their values. Those factors might be more important in driving financial and productivity gains than their coordinated commitment to AI.  

“The fact that organizations who adopt these more sophisticated AI uses have higher revenue or whatever cannot be disentangled from the fact that a wealthy organization can afford to buy stuff,” Boyd says.

Anderson acknowledges that wealth and size could be a factor, noting that the study doesn’t claim that AI maturity alone caused revenue and productivity growth.

“What our findings suggest, however, is that organization size alone does not fully explain the differences we’re seeing,” she says. “The strongest outcomes consistently aligned with AI maturity.”

Blackbaud, which both conducts research and sells AI and other technology products, measured charities’ use of any AI products, including widely used chatbots like ChatGBT and Claude in its report.

Nonetheless, Boyd says, any study of AI productivity gains by a company that sells AI products should be viewed with a healthy degree of skepticism.  

Blackbaud said in a statement that its research lab uses large-scale data on the sector along with primary research from partners to better understand trends in the nonprofit world. Anderson, the report’s main author, said she tries to “put up a little bit of a wall between myself and our products to keep that objectivity.”