Changing what works to what scales

Tech is shifting faster than the models we built our impact on. And that means even thriving nonprofits face a choice: Keep optimizing what works—or rebuild for what’s coming.

Back in June, our leadership team made a decision that felt both risky and obvious: Change a strategy that was still working to accommodate an AI future. We’d been writing and speaking for years about the need for the social sector to stop talking and start doing—and we realized it was time to take our own advice.

For the last five years, our organization has helped nonprofits worldwide build tech solutions in partnership with leading tech companies. It worked. It made a difference. But by 2025, it became clear: What brought us to this point won’t take us to where we want to go.

We could keep matching tech needs with builders. Or we could bet on something bigger—teach nonprofits how to prepare for an AI-native future, so they can be capable of building and scaling impact themselves. We chose the latter.

A BET ON THE FUTURE

In recent weeks, four major reports were released: The Philanthropic Reset, AI for Humanity, Accelerate What’s Possible, and AI With Purpose.  Four different sources, same message: Nonprofits are ready for AI—but the systems around them are not.

The data is clear:

• 84% of AI-powered nonprofits lack funding to further develop and scale AI solutions.
• 87% of funders admit they don’t understand their grantees’ tech capacity.
• 90% of nonprofits don’t fund AI literacy or infrastructure.

And yet, the organizations seeing the biggest results are those that fine-tune AI with their own data, test quickly, and integrate community feedback. The takeaway is simple but uncomfortable: The real bottleneck isn’t technology—it’s capacity.

That realization pushed us to rebuild not just our programs, but our mental model of what “tech for good” means in an AI-native world.

FROM ONE-OFFS TO ECOSYSTEM

For years, the social sector has measured success by the number of pilots launched. But in the AI era, pilots don’t scale. Systems do.

So, we’ve started building what we call an AI enablement ecosystem—a space where nonprofits can build, learn, and scale responsibly, together.

That includes initiatives that help organizations prototype their first AI tools and build internal capacity, support proven social solutions so they can scale through responsible AI use, and a venture-style lab that develops shared infrastructure for nonprofits.

But this isn’t about our model. It’s about a broader shift—from delivering solutions to building systems that deliver.

WHAT “AI-NATIVE” REALLY MEANS

Being AI-native doesn’t mean asking ChatGPT to write your next grant report. It means processes, interventions or even full organizations that make the most out of the promise and benefits of AI.

Imagine a three-person nonprofit running a program that today would require a staff of 30. AI handles logistics, data analysis, and reporting, while humans focus on relationships, trust and connection. That’s not that far away. It’s already happening.

And it forces us—leaders, funders, and builders—to rethink what kind of infrastructure we’re really financing. Are we funding innovation, or the capacity that makes innovation possible?

Our bet is simple: In the next two to three years, it will be exponentially easier for nonprofits to build and scale with AI. But for that to be safe and responsible, we’ll need a shared layer of infrastructure—capacity, governance, and collaboration that helps changemakers build with confidence.

We’ve spent years telling the sector to stop talking and start doing. This is why we’re doing it ourselves. Because in the end, doing the right thing isn’t about keeping what works. It’s about having the courage to rebuild while things still work. And that’s exactly what the moment demands and the technology enables.

Jacek Siadkowski is the founder and CEO of Tech to the Rescue.

View the full article