SoliContraX, October 9, 2026

How to Build Grant Skills That Last as AI Changes the Work

AI is changing how applications get drafted, and funders are responding. Here's where the grant profession is shifting and which skills to invest in.

If you've watched AI tools get better at drafting, you may wonder what your own role will look like in a few years. Worry is reasonable, and so is curiosity. Nobody can predict the profession precisely, but funders' own published rules tell you something about where the work is moving.

This article looks at what is changing, what funders are doing about it, what stays in human hands and how to pick one skill to build over the next twelve months.

What is changing in the work

Drafting is getting cheaper. A first version of a narrative, an organizational background or a budget justification can now start from a tool instead of a blank page. That changes how you spend a week: less time producing a first draft, more time deciding what the draft should say and whether it is right.

It also lowers the cost of producing applications in bulk, and at least one major funder has said so in writing. In its July 2025 notice NOT-OD-25-132, NIH says that AI tools have enabled some principal investigators to submit more than 40 distinct applications in a single submission round. NIH describes the percentage of investigators averaging more than six applications per year as relatively low. For applications received on or after September 25, 2025, it limits each principal investigator to six new, renewal, resubmission or revision applications per calendar year, and says it will not consider applications substantially developed by AI to be the original ideas of the applicant.

You don't need to apply to NIH to learn from this. When a tool makes volume easy, funders respond with limits, originality expectations and disclosure requests. Expect more of that, and expect the details to differ by funder. Funder AI policies shows how to find the rule that applies to you.

What stays with people

Several things don't move when drafting gets faster.

Responsibility. NSF states that proposers are responsible for the accuracy and authenticity of their proposal submission, including content developed with generative AI tools. NSF also encourages proposers to say in the project description whether and how they used it. A tool can't take that responsibility from you.
Reviewer protections. NIH prohibits its peer reviewers from using generative AI tools to analyze applications or write critiques, and NSF prohibits reviewers from uploading proposal content to tools it has not approved. Whatever happens on the applicant side, people are still doing the reading and scoring.
Truth. A draft that states false things is still a false application. Checking facts doesn't get automated away; if anything, there is more to check. Fact-checking AI drafts covers the routine.

What reviewers may see

Nobody has good public data on how AI is changing the applications that reviewers read, so treat this section as reasoning rather than measurement. If many applicants give a tool little beyond a notice and a one-line prompt, the drafts they get back will tend to resemble each other: the same stock phrases, the same general claims, the same tidy structure. A tool can't supply details it was never given.

For you, that points in one direction. Polish stops being a differentiator when polish is easy. What can still distinguish an application is what only your organization knows: the place, the people, the track record, the community's own words and a program design that matches the funder's stated priorities. Keeping your organization's voice covers how to put that into a draft.

Be open about how you work

Professional credibility will rest partly on how plainly you describe your process. Where a funder requires disclosure of AI use, disclose it, and where it only encourages disclosure, consider doing so in a sentence or two: what the tool was used for, what you did with the output and who is responsible for it. Where a funder limits AI use, follow the limit. Nobody should be advising you to hide how a document was produced, and it isn't a sound basis for a career if you do.

Being open also helps colleagues. When your team knows which tasks you hand to a tool and which you check by hand, new staff learn the practice quickly, and your board can see where the judgment sits.

Skills that matter more

Five skills grow in value as drafting gets cheaper. You can build each one on purpose.

Strategy. Deciding which funders to pursue, which to skip and what to ask for. A tool can summarize a notice; it can't know your capacity.
Program design. A strong proposal rests on a program that makes sense. If the design is weak, a smoother draft only exposes it more clearly.
Relationships. Program officers, partners, community members and board members all matter, and trust is built through conversations.
Editing and verification. Cutting, sharpening, adding the specific detail and confirming every claim. This is where drafts become yours.
Judgment about tools. Knowing what to hand off, what to keep, what a funder permits and what data you must never upload.

Say two development directors at similar nonprofits both start using a drafting tool. The first accepts what comes back with light edits. The second spends the saved time calling a program officer, interviewing a program lead for specifics and checking each number against the audit. Both finish earlier than before. Only the second's application contains something a stranger couldn't have written, and the second director has also built skills that travel to the next job.

How to build one skill on purpose

Choose the skill that is weakest for you or most valuable to your organization, and make it concrete.

Set a measurable practice. For relationships, that could be one funder conversation a month. For editing, it could be a specifics pass on every draft.
Find a source for feedback. Ask a colleague or a program officer to read one section and say what was unclear.
Review it at a set date. Put a check-in on your calendar six months out and note what has changed.

If your skill is the career path itself, the grant professional career path looks at the roles and the ways to move among them.

Your next step

Look at the five skills above and pick the one you most want to be known for in two years. Write one sentence describing a specific action you'll take in the next 30 days to build it, and put the date for it on your calendar.

SoliContraX is built on the idea that tools draft and professionals decide. If you want the first draft off your plate so more of your week goes to the skills above, upload a funding notice and start from a draft written from your own materials. Get started with SoliContraX.

Sources

All posts