AI as Your Starting Point, Not Your Ghostwriter

Let me tell you what an AI-generated proposal looks like.

It looks great.

The headers are in the right places. The structure mirrors the solicitation. The language is professional. The logic model terminology is correct. The sections flow. If you read it quickly, it feels like a real proposal.

And that is exactly the problem.

Because a funder's program officer does not read proposals quickly. They read them slowly, critically, and with over 17 years of experience reading proposals from organizations doing work, trust me…they know deeply. And what they notice, almost immediately, is the absence of the thing that makes proposals win.

They notice that no one is home.

What Program Officers Actually Notice

An AI-generated proposal tends to describe a type of organization doing a type of work for a type of beneficiary. It is technically correct and completely generic. It could have been written for any organization in any city doing roughly similar work.

A winning proposal describes this organization, in this community, doing this work for these specific people. It uses language that only an insider would use. It makes reference to local context that reveals genuine knowledge. It expresses a theory of change that is particular and is the result of actual learning from actual implementation.

That particularity is what AI cannot manufacture. And experienced reviewers feel its absence instantly.

Formatting and structure are absolutely important. But they are not the mechanism that makes a funder fund a project. What makes a funder fund a project is connection: to the heart first, and then to the head.

Where AI Belongs in Your Process

Use AI For: Research and Synthesis

We covered this in Newsletter 1. AI's synthesis superpower is real. Use it to build your literature base, understand your funding landscape, synthesize field-wide trends. This is where AI saves you the most time with the least risk.

Use AI For: Generating a Scaffold

Once you have completed your five pillars (recall that this is the deep pre-work that only you can do), AI can be genuinely useful for generating a structural scaffold for your narrative. Feed it the solicitation, your pillar answers, and ask it to create an outline that maps your content to the funder's requirements. Then you fill that scaffold with your voice.

Think of it like a building frame. AI gives you the steel structure. You supply the walls, the windows, the texture, the warmth. This is everything that makes it a place someone actually wants to enter.

Use AI For: First Drafts of Boilerplate Sections

Every proposal has sections that are more structural than expressive, for example, organizational history, fiscal capacity, compliance narratives, and so on. AI can produce solid first drafts of these sections that you then personalize and verify. This saves meaningful time on the parts of the proposal that matter least.

Use AI For: Editing and Refinement

Once you have written your narrative, AI is an excellent editing partner. It can check for clarity, flag passive voice, identify sections where your logic has gaps, and suggest tightening. This is AI in service of your writing and not AI replacing it.

Where AI Does Not Belong

Not For: Your Opening Narrative

The first page of your proposal is where a funder decides whether they are going to keep reading or set it aside. That page needs to be entirely, unmistakably you. No AI scaffold, no AI language. Write it yourself, from the deepest place you know.

Not For: Your Theory of Change

Your theory of change is why your activities produce your outcomes for your specific beneficiaries. This is the intellectual and experiential core of your proposal. It is your organization's wisdom made explicit. AI will give you a generic one. You need to write the real one.

Not For: Your Impact Stories

We talked about this in Newsletter 3. The heart of a proposal comes from your lived experience of transformation in your community. AI cannot access that experience. You cannot delegate it.

The One Question to Ask Yourself

Before you use AI for any part of your proposal, ask yourself: does this require my organizational expertise and voice, or is this a structural task that AI can accelerate?

If it requires your expertise and voice, write it yourself.

If it is structural, let AI start it, then make it yours.

That distinction, applied consistently, is the difference between an AI-assisted proposal that wins and an AI-generated proposal that doesn't.