Ask Your AI Who Funds Nonprofits Like Yours. Then Ask Where the Answer Came From.
More people are asking ChatGPT or Claude where to find grants. It is a reasonable thing to try. You type a question, and a few seconds later you have a list of foundations, a sentence about each one, and sometimes a suggested amount to ask for.
The trouble is that the list sounds equally confident whether it is right or not. Grant writers have started trading stories about it: a foundation that stopped giving years ago, a program that was renamed, a funder that does not exist at all. The assistant is not lying. It is doing what it was built to do, which is to write a plausible answer from what it has read. And very little of what it has read is the record of who actually gave money to whom.
Where the real answer lives
Every private foundation in the country files a tax return with the IRS, and that return lists its grants: who got the money, how much, and in what year. That is the closest thing there is to a complete record of foundation giving.
It is also a record almost nobody writes about. A foundation's grant list sits inside a tax filing, not on a web page, so an assistant searching the web mostly finds the foundations that publicize themselves. The quiet ones, the family foundations that write the same check every year and have never built a website, are close to invisible.
GrantSnag reads those filings and matches each grant back to the organization that received it. It also connects directly to Claude, ChatGPT and other AI assistants, so when you ask your assistant who funds a nonprofit, it answers from the filings instead of guessing.
One question, asked properly
We tried it on a nonprofit near us: Seton Youth Shelters, a youth shelter in Virginia Beach. We typed one sentence:
Who funds Seton Youth Shelters in Virginia Beach?
Here is what came back, all of it from filed returns covering 2020 through 2025:
- 20 foundations
- 61 grants
- $2,444,610 in total
A few of the details are worth slowing down on.
The largest funder is one you cannot apply to. The Agena Foundation in Chesapeake gave $1,067,500 across six grants, one in every year from 2020 to 2025. Its filing says it only gives to preselected organizations. An assistant guessing from the web might well tell you to send it a proposal. The filing tells you that would be wasted work, and that the way in, if there is one, is a relationship.
The second largest has a deadline, and it is printed in the return. BM Stanton Foundation in Virginia Beach gave $381,488 across six grants, again one every year. Its filing says requests come in before October 31 each year. That is the kind of fact that turns a name into a plan.
Most of the list is local. 14 of the 20 funders are in Virginia. Alongside the big names (the United Way of South Hampton Roads and the Hampton Roads Community Foundation) are small family foundations and local businesses' charitable arms.
More than half of them have no website. 11 of the 20 list no website in their filings. For those funders there is no page for an assistant to find. There is a tax return.
What it does not do
It is worth being plain about the limits, because they are real.
- It does not write your application. There is no AI inside GrantSnag. It hands your assistant facts; what you do with them is still your work.
- Filings lag. A foundation files after its year ends, so the newest grants in the record are usually a year or more old. A funder can change direction after its last return.
- Not every name is a grant program. A few entries on any list like this are giving platforms and donor-advised funds, such as Network for Good or National Philanthropic Trust. Those show you which of your donors routed a gift through them. They are not programs you can apply to, and a good list says so rather than padding the count.
- Preselected means preselected. Five of the 20 funders on Seton's list say in their filings that they only give to organizations they choose. That is useful to know, and it is also a closed door until someone opens it.
Why start from an organization like yours
The question we asked was not "which foundations fund youth shelters." It was "who funds this particular shelter." That backwards search is the one a web search cannot do, and it is the one that finds the quiet local funders. Run it on two or three organizations that look like yours. The funders they share, and that have not funded you yet, are your best prospects. We think it is the best peer lookup available, and it now sits inside the assistant you already use.
Trying it yourself
If you use Claude or ChatGPT, you can connect GrantSnag in a couple of minutes. The connection page has the address to add and walks through a full example, start to finish. It is included with every paid plan, and monthly plans start with a free trial.
Then ask your assistant who funds an organization like yours, and ask where every name came from. With the filings behind it, it can tell you.