Embedding AI is much more about people and culture and change management than it is about a specific tool or technique or best practice.

Remy ReyaRemy ReyaAI Expert, Nonprofit Cybersecurity/Tech Readiness Cohorts, NTEN

Nonprofit funders are still finding their footing with AI, and so are many of the organizations they support. Staff across the sector already use AI tools, but few nonprofits have had the guidance or resources to build those tools into how work gets done. Progress depends on foundations and grantees working through the technology together.

Remy Reya is Director of AI and Thought Leadership at Compass Pro Bono, a nonprofit that connects local business professionals with community nonprofits through pro bono consulting. Reya has spent more than a decade in the social sector, including several roles at Compass, and has trained over 3,000 people on values-aligned AI. A TEDx and keynote speaker, Reya presents regularly to nonprofit and philanthropy audiences.

"Embedding AI is much more about people and culture and change management than it is about a specific tool or technique or best practice," says Reya. Neither foundations nor grantees need to have AI figured out before they start that work.

Money shapes adoption

Only 4% of nonprofits say they've fully integrated AI into their operations, according to a 2026 sector survey from the nonprofits Bridgespan and NTEN. Most foundations haven't set expectations for how grantees use AI either. A 2025 research report from the Center for Effective Philanthropy, a nonprofit research organization, found that 98% of foundations have no policy on whether grantees can use AI in their work.

That gap matters because nonprofits tend to invest in the areas their funders encourage and pay for. "If foundations are saying nothing about AI and aren't giving their grantees guidance on AI, it's understandable that nonprofits may not even have the resources to invest in training or upskilling or the change management work that is so critical to meeting this moment," says Reya.

Funders are only one of the groups nonprofits answer to, alongside corporate donors and the communities they serve. AI carries different risks for each of those groups, and nonprofits have to weigh all of them at once.

Paying for safety

Nonprofits are better off paying for business versions of their AI tools. Free versions have weaker security protections, and leaders can't see or track what staff do in personal accounts. "If you're not paying, you are the product. In the case of AI, it's your data. Your data is being retrained into the models, and that creates this new surface area of risk," says Reya.

Staff often use AI even when their organization tells them not to, so it's safer to give them an approved tool. Nonprofit discount programs keep many of these plans affordable, and some are free. To choose a vendor, leaders can compare an AI tool's terms with those of the platforms that already hold their sensitive data, such as Microsoft 365, Google Drive, or Salesforce. Many AI vendors are startups that update their terms often, so that comparison needs repeating. "If you can find parallel clauses in the terms and conditions for an AI tool, great. That gives you a starting point. It doesn't need to be the ceiling, but it gives you a floor for conversation," notes Reya.

Principles over policy

Clear principles are easier to keep current than a formal AI policy, which can take months to approve. "You can't go through a multilayer, 10-month-long policy approval process for AI because it's going to be outdated by month three," says Reya.

Those principles work best when they name the specific concern behind a rule. A leader who wants to ban a certain tool, for example, may be most worried about staff letting AI make critical decisions. A principle about which decisions stay with people holds up no matter which tool staff use.

Principles still need people to apply them. Most teams have a few staff members who are curious about AI, and leaders can name them as champions who follow new tools and keep the conversation going. Giving them that role shows that leadership knows staff use AI and takes its effect on the mission seriously. "If we figure out what the baseline principles are that preserve our mission and the integrity of the work that we do day to day, they're going to be much more robust to the change in the AI space and allow us to take a deep breath instead of trying to run after every new update," says Reya.