AI Agents in Grantmaking: Are You Innovating—Or Just Paving the Cow Path? (Part 2)

Image Created by AI

This is Part 2 in a series exploring how the seemingly intuitive path—automating familiar processes with AI—may not be the most effective application of these powerful tools. For grantmaking foundations, this pattern can be especially limiting—and in some cases, counterproductive.

As AI continues to mature, it’s easy to get swept up in the excitement. AI agents are being pitched as digital staff—handling everything from application review to grantee communications and reporting. The promise? Lower costs, faster decisions, and round-the-clock productivity.

But before we jump in, let’s pause to ask the critical question: Are we truly innovating—or just paving the cow path?

The Risk of Automating What’s Broken

Most grantmaking systems today are digital mirrors of paper-based processes—forms turned into web portals, manual reviews encoded into scoring rubrics, reporting forms made clickable.

Sure, automating these tasks may speed things up. But faster isn't always better—especially if you're simply accelerating processes that were never designed for the outcomes your foundation values today.

This is the classic “pave the cow path” trap: using cutting-edge tools to optimize outdated workflows, instead of asking whether the workflow itself still makes sense.

A Parallel in History 

In the early railroad era, tunnel construction was done by hundreds of laborers wielding pickaxes and explosives. Progress was slow and dangerous. Imagine trying to modernize that operation by replacing every worker with a humanoid robot holding a pickaxe.

Sounds futuristic—but it’s still incredibly inefficient.

The real leap came when engineers stopped trying to mimic the old process and instead reimagined it entirely. Enter the tunnel-boring machine: a massive, fully automated system guided by a small team of technicians. It didn’t just speed up the work—it redefined how the work was done.

The lesson? Don’t automate each step. Redesign the system.

From Automation to Transformation

In grantmaking, we should do the same. Don’t start with the workflow—start with the mission. Reexamine what your foundation exists to do. What impact do you want to have? What principles should guide your decisions?

Only then should you look at your grantmaking process and ask:

• Where does AI add real value?

• Where is human judgment, empathy, or relationship-building irreplaceable?

• What decisions could be improved with better insights, not faster processing?

Used well, AI can be a powerful co-pilot. It can help uncover hidden trends, identify funding gaps, flag bias in decision-making, or streamline internal operations so staff can focus on strategic work.

But AI should not be a digital impersonation of a grant administrator or a program officer. It should be a tool that amplifies your team’s thinking, not replaces it.

Reimagine First. Then Automate.

Let’s be clear: the real opportunity here isn’t automation—it’s reimagination.

Don’t retrofit AI into workflows that were never designed for equity, flexibility, or learning. This is your moment to rethink how you work, and rebuild your processes around what truly matters: your mission, your grantees, and the outcomes you seek in the world.

Let’s not pave the cow path. Let’s build a better road entirely.