If you’ve ever spent hours crafting a grant report that you suspect no one will read, you’re not alone. Written grant reports are the paperwork everyone loves to hate, both time-consuming for nonprofits and often left unread by funders.
The Houston Endowment decided to try something different. In a pilot program launched last year, they ditched written reports for their civic engagement grantees and replaced them with recorded Zoom conversations. Program officers and grantees talked through progress, challenges, and outcomes, and then used AI tools to transcribe and synthesize the discussion. The result was richer information, less busywork, and actual conversations between foundation and nonprofit staff.
What Houston Endowment Actually Did
Instead of requiring grantees to fill out written responses through an online portal (a process that can take several hours, especially if you’ve ever experienced the dreaded logging back in and your previous session hasn’t been saved) the Houston Endowment scheduled 45- to 90-minute virtual meetings. Grantees received discussion questions in advance and gave permission to be recorded. Zoom’s AI tools transcribed the conversations, and staff reviewed the transcripts for accuracy.
The shift wasn’t just about saving time. According to Robiel Abraha, a learning and evaluations officer at the foundation, oral reports captured nuance that written reports miss. Follow-up questions, storytelling, and context are all important parts of reporting, but are often the kinds of things that get lost when you’re trying to summarize a year’s work into text boxes.
Miguel Rivera, a program officer for the endowment’s civic engagement program, noted that these conversations revealed trends across the field more effectively than stacks of written reports ever did. When you’re actually talking with grantees (rather than just reading their reports) and using AI to generate and review summaries, patterns emerge.
The pilot worked well enough that the endowment is now expanding it to other programs, including arts and culture.
Why This Matters (And Where It Gets Complicated)
Replacing several hours of report-writing with a 60-minute conversation is a gift to already overstretched nonprofit staff. And let’s be honest about what often happens to those written reports: they sit unread, or they get skimmed for compliance purposes, or they become part of a board packet that no one fully digests.
As a former Program Officer, I often engaged in conversations with nonprofit staff, but would have valued these AI tools to help distill information and identify patterns in our region and in our work. Conversations help surface the real story about what’s working, what’s not, and where an organization is struggling. Written reports tend to be carefully managed narratives designed to make everything sound fine. Oral reports, when done well, invite honesty.
There’s also the hidden cost issue. When a nonprofit spends 10-15 hours on a grant report, that’s 10-15 hours of funded work that didn’t happen. It’s essentially a reduction in the value of the grant. Streamlining reporting isn’t just about convenience, it’s also about maximizing impact.
AI’s role in identifying trends is particularly interesting. Program officers can start to see patterns across multiple grantees without manually combing through dozens of documents. That kind of field-level insight is valuable for strategic grantmaking.
But this approach isn’t without complications. Trusting relationships matter here. Grantees in the Houston pilot were initially concerned about whether AI could capture “the essence” of their work or whether something would get lost in translation. The endowment’s existing relationships with these organizations made the transition smoother. If you don’t have that foundation of trust, asking grantees to be recorded might feel invasive rather than collaborative.
There’s also an equity question here. The Houston Endowment’s approach works because they’re providing additional capacity building support, including funding through their Collaboration Fund to help cover technology costs. Not every foundation is doing that. If oral reporting becomes the norm but only well-resourced organizations can participate effectively, we’ve just created a new barrier.
Also, let’s not ignore the broader hesitation around AI in the nonprofit sector. A recent Center for Effective Philanthropy report found that while some AI use is happening (drafting emails, that sort of thing), many foundation and nonprofit leaders don’t fully understand the technology. Concerns about accuracy, privacy, bias, and lack of expertise are real. More than 80 percent of foundation leaders worry (rightfully so) about data and privacy risks.
What You Should Do With This Information
If you’re a foundation, pay attention to what Houston Endowment is testing. You don’t need to adopt their exact approach, but you should be asking whether your current reporting requirements are actually useful or just performative. Are you getting information that improves your grantmaking, or are you collecting documentation to check a box?
Consider whether oral reporting—with or without AI—might work for at least some of your portfolio. Start small, build trust, and be transparent about how you’ll use the information.
If you’re a nonprofit, there’s an opportunity here beyond waiting for your funders to change. AI tools can help you synthesize your own work. You can “interview” yourself about a project, either by typing into a tool or using voice-to-text features, and have AI help organize themes, identify trends, or draft sections of a report. It won’t replace human judgment, but it can reduce the struggle of staring at a blank page.
The key is experimentation. Grantees in the Houston pilot, like the League of Women Voters of Houston and Mi Familia en Acción, are already exploring how AI might streamline their own operations, from chatbots for voter information to reviewing best practices for AI use.
This is still the early days of AI technology in the nonprofit space. AI in grant reporting isn’t a magic fix, and it won’t work everywhere. But if it means fewer hours spent on reports that no one reads and more time spent on actual conversation and learning, that’s a shift worth watching.
This post was inspired by reporting from Stephanie Beasley in The Chronicle of Philanthropy.
Image by Gerd Altmann from Pixabay.