Grant writing is the bane of every small nonprofit’s existence. It’s time-consuming, repetitive, and the success rate is depressingly low. A typical Australian charity might spend 40 hours on a grant application with a 15% chance of success. When your team is already stretched, that’s a painful investment.

AI tools — particularly large language models like ChatGPT and Claude — can dramatically reduce the time spent on grant applications. Not by writing the application for you, but by accelerating the parts of the process that are most time-consuming. Here’s how.

What AI is good at in grant writing

Drafting and structuring responses. If you’ve answered a grant question a dozen times before, you know the key points you need to make. AI can take those key points and draft a well-structured response in minutes. You then edit, refine, and add the specific details that make it compelling.

This isn’t cheating. It’s the same process as starting from a previous application and editing it, which every experienced grant writer already does. AI just does the first draft faster.

Summarising and condensing. Many grant applications have strict word limits. AI is excellent at taking a longer piece of text and condensing it to fit a word count while preserving the key messages. This is particularly useful for executive summaries and project descriptions.

Tailoring language to different funders. Government grants expect formal, structured language. Philanthropic foundations often prefer a warmer, more narrative style. AI can adjust the tone and register of your content to match different funder expectations, starting from the same core information.

Generating logic models and theories of change. If you give an AI tool your program description and intended outcomes, it can generate a draft logic model or theory of change framework that you can then refine. This isn’t a substitute for genuine program design, but it can accelerate the documentation process.

What AI is bad at

Knowing your organisation. AI doesn’t know your specific context, your community, your track record, or the nuances of your approach. Any AI-generated content needs to be reviewed by someone who deeply understands the organisation and can add the specific details, examples, and evidence that make an application compelling.

Understanding funder priorities. AI can read a grant guideline, but it can’t pick up on the unwritten priorities and preferences of a particular funder. An experienced grant writer knows that a certain foundation cares deeply about community engagement, or that a government department is prioritising a specific policy area. That contextual knowledge needs to come from you.

Being authentic. The best grant applications feel genuine. They convey passion, experience, and deep understanding of the problem being addressed. AI can produce polished prose, but it can’t produce authenticity. If your application reads like it was generated by a machine, experienced assessors will notice.

Handling complex budgets. AI can help with budget narratives, but the actual budget numbers need to come from your financial team. AI doesn’t know your salary costs, your overhead rate, or the real cost of delivering your programs.

A practical workflow

Here’s a workflow I’ve seen work well for small nonprofits.

Step 1: Read the guidelines thoroughly. No AI tool can substitute for carefully understanding what the funder is asking for. Read the guidelines, the assessment criteria, and any available information about funder priorities.

Step 2: Create a brief for each question. For each question in the application, write a brief dot-point outline of what you want to say. Include specific examples, data points, and evidence you want to reference.

Step 3: Use AI to draft responses. Feed your brief and the question into an AI tool and ask it to draft a response. Be specific about word count, tone, and any particular points you want emphasised.

Step 4: Edit rigorously. This is the critical step. Go through every AI-drafted response and edit it. Add specific details the AI couldn’t know. Remove anything that sounds generic or formulaic. Ensure the voice matches your organisation’s authentic communication style.

Step 5: Get a human review. Have someone who knows the organisation read the final application. Does it sound like you? Does it accurately represent your work? Would you be proud to put your name on it?

The ethics question

Some nonprofits worry about whether using AI for grant writing is ethical. Here’s my view: funders care about what you’re going to do with their money, not how you wrote the application. If AI helps you communicate your genuine plans more clearly and efficiently, that’s a positive outcome for everyone.

The ethical line is dishonesty. If you’re using AI to fabricate evidence, inflate outcomes, or misrepresent your organisation’s capabilities, that’s wrong regardless of whether you used AI or not. The tool doesn’t change the ethics; the intent does.

Tools worth trying

For Australian nonprofits on a budget, ChatGPT (free tier) or Claude (free tier) are perfectly adequate for grant writing assistance. If you want more structured support, some AI consulting firms in Sydney offer workshops specifically for nonprofit teams learning to use AI tools effectively.

The key is practice. Your first AI-assisted grant application will probably take almost as long as writing from scratch, because you’re learning the workflow. By your fifth, you’ll have cut your drafting time in half.

The bottom line

AI won’t write your grants for you. But it can significantly reduce the time you spend on the parts of grant writing that are most repetitive and least creative. For time-poor nonprofits competing for limited funding, that time saving is genuinely valuable.

Just remember: the AI writes the first draft. You make it real.