# Why AI gives generic answers about your nonprofit

When AI doesn't know your organization, it guesses. Tell it about your world before you ask it for anything, and the generic answers start to disappear.

- Author: [Scott Williams](https://ifpossible.ai/about#scott), Co-Founder, If Possible
- Published: 2026-10-14
- Topic: How to Implement AI at Your Nonprofit
- URL: https://ifpossible.ai/articles/why-ai-gives-generic-answers
- Video: [Why Does AI Give Generic Answers About Your Nonprofit?](https://www.youtube.com/watch?v=zSU_bwmF9YQ) (5 min)
- Publication: [The Art of the Possible](https://ifpossible.ai/articles/), by If Possible (https://ifpossible.ai)

## Key takeaways

- AI gives generic answers because it's missing context about your nonprofit, and it fills every gap with a confident guess.
- AI gets context in three ways. You tell it in the prompt, it reads it from a brain you built, or it makes it up.
- For your first stretch with AI, tell it things more than you ask it for things.
- Great context beats a great prompt, because context keeps paying off on every task after.
- When a draft looks wrong, look for what you forgot to tell it before you blame the tool.

## Why does AI give generic answers about my nonprofit?

AI gives generic answers because it doesn't know your organization, and it never stops to tell you so. It fills every missing detail with a guess drawn from the average of everything it has read. The result sounds smart and could belong to any nonprofit in the country.

I explained this to a room of nonprofit leaders during a masterclass in October with a picture of buckets. Say a good story needs a hundred buckets of information filled in, like who the person is, where they come from, what they're up against and what they want to overcome. Tell a person a story with seven of those buckets filled and they walk away confused. Hand the same seven to AI and something different happens.

> AI will fill in the 93. It will, and then that's what makes it sound like it's really smart.
>
> Scott Williams · Recorded conversation, Oct 9, 2026

The question I asked that room is the one I'd ask you. Is it the right 93? When AI guesses, it guesses toward a vanilla blend of every nonprofit on the internet. Your donors, your programs and your voice are exactly the parts it can't guess.

## Where does AI get its context?

AI gets context from three places. You tell it in the prompt, it reads what you've stored in a brain or database, or it invents it. If you skip the first two, the third happens on its own, and [nothing on the screen warns you that it did](https://ifpossible.ai/articles/check-ai-work-for-errors).

I put it in one breath when Colleen and I were talking this through in September:

> There's three ways that AI is going to get context. It's either from what you tell it in the prompt, from what you tell it that lives in a brain or a database, or it will make up context.
>
> Scott Williams · Recorded conversation, Sep 10, 2026

Most teams live in the first place. They type a request, add a sentence of background and hope. That works for small tasks. It falls apart on anything that depends on knowing you, like an appeal, a board report or a grant narrative.

The second place is where the real gains are. At If Possible [we call it a mission brain](https://ifpossible.ai/articles/what-is-a-mission-brain), one organized home for everything about your organization that AI reads before it starts any task. Once it exists, you stop explaining yourself from scratch in every chat.

| Where context comes from | What it looks like | What you get back |
| --- | --- | --- |
| The prompt | A sentence or two typed into a chat | Fine for small tasks, thin on anything about your mission |
| A brain you built | Organized documents AI reads before every task | Work that sounds like your organization |
| AI makes it up | Nothing, because nobody gave it anything | A confident guess at an average nonprofit |

## Why doesn't AI remember what I told it last week?

Most AI tools hold only a little memory, and when it runs out they keep going without mentioning it. That's why the same tool can feel brilliant on Monday and generic on Thursday. The fix is a place outside the chat where your context always lives.

When I teach this, I compare AI to a goldfish:

> You're dealing with a goldfish. That's what AI is. But what you need to do is say, hey, that goldfish has enough connectivity to connect me to something that will always remember. And that is the brain.
>
> Scott Williams · Recorded conversation, Sep 14, 2026

I got caught by this myself in October, in front of a room of nonprofit teams at our Intensive. A video I'd made came out badly, and I was ready to lose my mind. Then I figured it out. I had switched accounts, and everything I'd taught the old one stayed behind. I thought I'd brought the right information over, and I hadn't. I gave it the five things I'd forgotten, walked away, and came back 20 minutes later to a video that was right on the first try. Same tool, same request. The only change was context.

## Should I tell AI things before I ask it for things?

Yes. For your first few weeks, spend more time telling AI about your organization than asking it to produce anything. It sounds backwards. It's also the habit that most changes the quality of everything AI makes for you afterward.

Here's how I said it to Colleen:

> don't ask anything of AI for a while. Only you tell it things for a while.
>
> Scott Williams · Recorded conversation, Sep 10, 2026

Think about a new hire. For a good while, you're paying them to learn. Nobody expects great work in week one from someone who has never seen your donor list. AI knows everything on the internet and nothing about your world, and it needs the same onboarding.

Colleen has the shortest version of this I know:

> I think context is everything as it relates to using AI well. And so the better your context, the better your output, full stop.
>
> Colleen Cook · Recorded conversation, Sep 8, 2026

## What context should I give AI about my nonprofit?

Start with what a smart new hire would need in their first week. Who you serve, what you do, how you sound, who gives and why, and what you're working toward this year. Rough notes are fine, as long as they live in one place AI can read every time.

Use this checklist to build the first version of your brain:

1. **Your mission in your own words.** The official version, plus the way your executive director says it out loud.
2. **Who you serve.** Two or three real stories, with names changed, that show the people behind your work.
3. **Your voice.** Three to five pieces you're proud of, like an appeal, a thank-you letter and a newsletter.
4. **Your donors.** Who gives, why they give and what they've heard from you lately.
5. **Your programs and results.** Only the numbers you'd stand behind in front of your board.
6. **This year's goals and limits.** What you're raising for, what you can't say, what's off limits.
7. **Your website.** Colleen told a room of nonprofit teams in October that a website is a wonderful document for AI to learn who you are.

Then, before any big task, paste this in:

``` Before you write anything, read everything I've shared about our organization. Then ask me up to five questions about anything you would otherwise have to guess. Wait for my answers before you start the draft. ```

A [brand voice system](https://ifpossible.ai/articles/first-ai-system-brand-voice) is a strong first piece of the brain, and the [nonprofit brand-voice prompt](https://ifpossible.ai/articles/nonprofit-brand-voice-prompt) gets you most of the way there in an afternoon.

## How do I know when bad output is a context problem?

Assume it is until you've checked. When a draft comes back generic, off-tone or wrong about your programs, ask what AI didn't know before you rewrite the prompt or switch tools. Then add the missing piece to your brain so you never have to say it again.

This is also why I care more about context than clever prompts. I told a room of nonprofit teams in October that great context is so much greater than a really good prompt, because it compounds. A clever prompt helps one task. Every fix you add to the brain helps every task that comes after it.

If you're just getting started, [your first 30 days with AI](https://ifpossible.ai/articles/first-30-days-with-ai) puts the brain in week one, and [treating AI like a new hire](https://ifpossible.ai/articles/treat-ai-like-a-new-hire) explains the onboarding mindset behind all of this. You'll find more in our [implementation articles](https://ifpossible.ai/articles/topic/implementation).

Want one practical idea like this each week? Join [the If Possible newsletter](https://ifpossible.ai/newsletter). And if you'd like help building your mission brain alongside other nonprofit teams, save your spot on [the cohort waitlist](https://ifpossible.ai/cohort).

## Frequently asked questions

### Why does ChatGPT give generic answers about my nonprofit?

Because it doesn't know your organization, so it fills every gap with a guess based on what an average nonprofit might say. Give it your mission, your stories, your voice and your goals before you ask for anything, and the answers start to sound like you.

### What is context in AI?

Context is everything AI knows about you and the task before it starts working. It comes from what you type in the prompt, from documents it reads in a brain or database, or from guesses it makes when you haven't told it enough.

### Is a better prompt the fix for generic AI output?

A better prompt helps one task. Better context helps every task. Most teams get further by building one organized home for their mission, voice and donor information than by searching for the perfect prompt.

### How long does it take to give AI enough context?

Plan on a few weeks of telling AI about your organization before you expect great work, the same way you'd onboard a new hire. You'll see better drafts within days, and the results keep improving as the brain grows.

## About the author

[Scott Williams](https://ifpossible.ai/about#scott), Co-Founder, If Possible. Scott helps ambitious nonprofit leaders build AI-powered operations, drawing on a decade of fundraising work across 100+ organizations.