How to Use AI in Grant Writing
(Without Losing Your Voice, Vision, or Integrity)
By Mary Keane, CyberWyoming (written with the help of AI)
Forwarding note by CyberWyoming: Never put anything into an AI tool that you consider proprietary or sensitive, unless you have thoroughly checked the tool’s security and privacy policy and feel comfortable. Often AI tools use your inputs for teaching their large language learning model and some in the security industry had dubbed these models large language liability models for that reason.
Artificial intelligence applications like ChatGPT and Perplexity are seemingly everywhere. As you may have already found, AI is good at synthesizing data, writing reports, and writing blogs (like this one, which had help from ChatGPT).
AI is also changing grant writing, but not in the way some people think.
AI will not magically win funding (don’t we all wish it would!). It won’t invent a compelling program, understand our communities, or intuit what a funder really wants.
What it can do is simplify some of the tasks: organize information, summarize requirements, and help us move faster through the most tedious parts of the process.
Used well, AI can be a useful assistant. Used carelessly, it can remove the shine from your story, introduce subtle errors, and make your proposal sound like every other proposal in the pile.
Most importantly, AI cannot create insight or judgement. It can pull together information, but it doesn’t have any insight into how our organizations shine at finding inventive ways to help people.
Here’s how nonprofits and small businesses can use AI wisely, and where to be cautious.
Where AI Is Useful in Grant Writing
1. Pulling information together quickly
All of us who have written grants know that pulling together the requirements can be difficult. The information can be scattered or buried in long funding announcements.
Compliance language can be dense. Multiple attachments and forms can be hard to find, and sometimes the RFP will use a form name but the name of the form is different on the website. Occasionally deadlines or requirements even conflict with one another. It’s enough to make a grant writer need additional cups of tea and bars of chocolate.
AI can help with much of these issues. We have found it to be good at summarizing the RFPs and grant guidelines. It can even create a list with checkboxes to ensure we have everything completed before submission.
It can help with extracting requirements and attachments from public websites, although it cannot access password-protected portals. In that case, we have to download the pages as PDFs and feed them to the AI.
An example of what AI can do well:
Paste a funding announcement into AI and ask it to list:
Eligibility requirements
- Required proposal sections
- Mandatory vs. optional attachments
- Deadlines and formatting rules
This alone can save hours and reduce costly mistakes.
2. Creating first-pass drafts from our material
AI is very good at turning our notes and our websites into a draft proposal.
We have used it for drafting:
- Organizational background sections by pulling from our website
- Program descriptions from the website and our notes
- Compliance language
- Plain-language summaries from our notes and previous grant applications
Notes + website → AI → First Draft Proposal
AI drafts are more like raw clay, not finished work. The draft will be competent, clear, and structurally sound, but not memorable.
3. Editing for clarity and consistency
AI can help you:
Why Efficiency Matters
AI can help with tedious tasks because the stakes are high.
Given the high volume of applications and the rise of pre-screening letters of inquiry, even small efficiencies in grant writing can make a difference.
Caution: Do Not Share Confidential or Client Information
AI tools should never be used with confidential, sensitive, or client-specific information.
This includes:
Even when AI tools claim not to “store” data, we should assume that anything we enter could be logged, reviewed, or used for system improvement. That risk is not theoretical: it’s real, and it can create serious ethical and legal issues for nonprofits and small businesses.
A good rule of thumb:
If you wouldn’t paste it into a public document or send it to a stranger, don’t put it into AI.
Instead, use AI with:
AI can help shape structure, language, and clarity—but responsibility for confidentiality always remains with us.
Where AI Falls Short and Why That Matters
1. AI is not creative (especially in grants)
This is the most important caution.
AI does not create new ideas. It recombines existing language based on patterns.
In grant writing, that means:
- Our stories that make our organizations and missions so important are missing
- Safe, generic phrasing (the same phrases that everyone uses)
- Bland need statements (remember, AI doesn’t understand need; it just understands patterns)
- Over-polished and emotionally flat stories and examples (AI can bleed all of the drama and urgency out of a story)
Grant reviewers read hundreds of proposals that sound perfectly reasonable and completely forgettable.
What gets funded?
- A sharp insight
- A specific local need
- A bold or unusual approach
- A clear sense of urgency
- A strong ethical or community-centered stance
AI cannot invent these for the grant writer. These come from human judgement, lived experience, our mission, and our values.
2. AI does not understand context or politics
AI cannot:
- Read between the lines of a funder’s priorities
- Understand local relationships or history
- Know if similar proposals have been rejected by this funder several times
- Sense when language might alienate or resonate
- Know when to push and when to comply
AI writes toward safety and familiarity. When a proposal carries political risk, challenges existing systems, or ventures into untested territory, AI instinctively neutralizes it. The result is language that may be polished, but lifeless — stripped of the joie de vivre, urgency, and moral clarity that often persuade funders to say yes.
Human discernment is essential here.
3. AI hallucinations are real and dangerous in grants
AI may:
- Invent statistics
- Attribute quotes incorrectly
- Confuse agencies or programs
- State requirements confidently and wrongly
This is especially risky because AI sounds authoritative even when it’s wrong.
Rule of thumb:
Never submit a statistic, citation, eligibility claim, link, or compliance statement you did not personally verify.
AI is excellent at sounding right. That’s exactly why you must double-check everything.
This is just the beginning of the conversation.
In the next part of this series, we’ll dive deeper into practical tips for integrating AI into your grant writing workflow, including best practices, an AI Prompt Toolkit, and grant resources. Stay tuned!
