How AI is a Gamechanger for Nonprofits

Like many, I am deeply troubled by much of the rhetoric coming out of the AI labs, and the very real impact that AI is having (and will have) on our future. I am neither a “Doomer” nor a “Boomer” – more of an apocaloptimist (thanks, AI Doc) who feels strongly that there needs to be a new approach to AI, including much more safety testing, increased transparency and accountability, and some amount of government oversight. When over 1,300 employees from leading AI labs sign a letter suggesting a slowdown in AI research and more focus on safety, we should pay attention. Undoubtedly, cognitive offloading is real and will impact generations of learners. The explosion of data center construction is also deeply troubling, and I’m happy to see that many communities are pushing back, even when they need the revenue.

Recently in the New York Times, a former Lululemon CIO cautioned readers to be very skeptical about the impact of AI.  While it was a compelling piece, her perspective was exclusively around for-profit companies, which have one goal: increase shareholder value. So, yes, if that’s the goal, AI can be a finicky partner. While it might seem to bring much-needed efficiencies, the difficulty in properly leveraging AI, combined with the impact on jobs, makes it a net negative in her view.

However, including nonprofits in the conversation changes the equation significantly. Nonprofits are very adept at finding ways to do more with less. It’s in their DNA. So, if AI makes them 5% more efficient, they’ll take it, and they’ll “reinvest” in the mission. And since nonprofits are human-centered, the risk of AI taking over is much smaller. Moreover, while there could be some impact on jobs, mass layoffs don’t happen very often. The nonprofit sector needs more people, not fewer. But what about the tension between the positive impact on efficiency and the risk of replacing human connection in work that’s built on relationships? This is often referred to as dialectical thinking; as Fitzgerald said, “The test of a first-rate intelligence is the ability to hold two opposed ideas in mind at the same time and still retain the ability to function.”

Holding seemingly conflicting truths together is a well-understood place for nonprofit professionals; our field leverages the results of capitalism for the greater good. We’re very good at holding two opposing truths, which makes the AI conversation a familiar place for us.

It’s impossible and irresponsible to look past the impact this tool can have on this critical sector. As funding for nonprofits gets harder to secure, anything that can be done to strengthen the sector is worth considering. At the Jim Joseph Foundation, we take a measured but committed approach: use AI when appropriate, be transparent about its use, ensure that humans are driving, regularly evaluate the output, engage stakeholders, and continue to understand and review the impacts of this technology on your work. In that vein, we support our grantees to leverage AI by educating them on the possibilities and supporting them in its use. Our most recent initiative is piloting the concept of an AI Lead role inside a select group of organizations we support – someone responsible for determining where AI makes sense and ensuring it is used responsibly. While our learnings are just beginning, here are two examples where AI is making a difference today:

At BBYO, AI is embedded in how the IT team now serves the organization. Claude has deep context about BBYO’s systems, processes, documentation and even the way the organization is structured. That means technology staff can describe a problem in the language they use every day and ask AI to help diagnose it, propose solutions and, in some cases, build the code needed to fix it. When families were having trouble logging into BBYO’s online portal, for example, Claude helped the team identify what was going wrong, suggested changes to the login process and customer-service guidance, and wrote code that BBYO could test and refine. BBYO has since used the same approach to make it dramatically easier to add teens and families who attend events into its database and to address the enormously complicated task of assigning teens to the correct regions based on constantly changing ZIP codes.

But perhaps the most important impact is on people. A more junior Salesforce administrator at BBYO is now able to take on work that previously would have required considerably more experience. Rather than replacing her expertise, Claude is helping her develop it—functioning almost like a “digital mentor.” BBYO is seeing something similar with a web design specialist who had little Salesforce experience but is now using AI to bridge that knowledge gap and build within the platform. AI is helping the BBYO IT team become more capable, confident and ambitious in the work they can take on.

At UpStart, AI is rooted in a familiar Jewish idea: na’aseh v’nishma— “first we will do, and then we will understand.” Rather than waiting for the technology to mature or developing a perfect AI strategy, UpStart encouraged its staff to start experimenting and learn through use. Importantly, the organization framed AI as professional development, not human replacement. Some initial skeptics became among UpStart’s heaviest AI users, and staff in programs, finance, HR and operations were given additional resources to experiment within their functions. The key wasn’t forcing AI on people; it was letting colleagues see how useful it could be in someone else’s work and then wanting to try it themselves.

The results are already changing how work gets done. One employee had been spending roughly six hours each week gathering news from across UpStart’s network of 250 organizations to share with board members. UpStart built an AI-supported process that now gathers much of that information and prepares a draft for her to review and refine. More important than saving six hours was the fact that now she had more time to talk with board members, build relationships, and make connections.

AI also allows UpStart to think differently about scale and its goal to deliver five times the impact without increasing its budget fivefold. UpStart is now developing an AI version of its Jewish social innovation and entrepreneurship methodology that could eventually make its expertise available to thousands rather than hundreds of people each year. It is also testing AI “sidecars” that program participants can use between sessions to revisit content, ask questions and work through challenges. The goal isn’t to replace coaches, facilitators or relationships, but to make UpStart’s knowledge and support available more often and to more people.

These are just two examples of our grantee-partners embracing concerns about the potential negative impacts of AI while also pursuing its usage to achieve greater impact. This dialectical thinking reflects a long-standing Jewish teaching. In Eruvin 13b of the Babylonian Talmud, the schools of Hillel and Shammai argued for three years, each insisting the law should follow its view. Then a heavenly voice declared: “Elu v’elu divrei Elohim chayim”, meaning “These and these are the words of the living God.” But the voice kept going: “v’halacha k’Beit Hillel”, meaning “and the law follows Beit Hillel.”

So the teaching has two parts. Both positions carry truth, and a decision still gets made.

The Talmud then asks the obvious question: if both are the words of God, why does Hillel’s school win? It isn’t because they were more correct. It’s because they were kind and humble. The Rabbis studied Shammai’s arguments alongside their own, and they even stated Shammai’s position before their own when teaching.

Importantly, the tradition also preserved the minority’s view. Shammai’s arguments stay on the page, and the Mishnah explains that they’re kept so a later court could draw on them if circumstances change.

If we apply this to the topic of leveraging AI in our work in Jewish nonprofits, it guides us to consider both truths, even if they’re in conflict. It teaches us to be humble and open to all arguments, but then decide how best to move forward, while remembering to document our decisions and revisit them as needed.

This is by no means the end of the conversation. AI is happening at lightning speed, so we have to continue to be deeply informed and engaged. Ensuring we maintain a deep connection to our humanity is becoming more critical and getting harder. Perhaps there will be events that change our approach to AI, or perhaps we’ll continue to operate in the space between. It is hard to predict. But, if you’re interested in this discussion, we’re here to engage.

David Goodman is Program Director, Technology Strategy at the Jim Joseph Foundation.