The distinction between a good grant application and one that just sounds good matters. However, most people can’t tell the difference.
This concept is particularly important to grasp now that artificial intelligence LLM tools have made it easy to quickly generate polished, professional-sounding writing of all types. AI can absolutely produce a grant application that uses sophisticated vocabulary, maintains consistent formatting, and creates sentences that seem to flow smoothly from one to the next. To an untrained eye, it looks impressive. On a first quick sweep of the form, even professionals like myself are impressed by the apparent quality. However, more often than not, this quality is just an illusion.
What might happen if you asked a ‘regular’ person to evaluate a bottle of Château Margaux? Sure, they might appreciate that it looks expensive. They may point out the elegant label, and see that the liquid is clear and richly coloured. They might enjoy the taste and even be able to pick out what flavours are in it, but could they explain what sets the flavour apart from a decent supermarket alternative? Could they tell you with confidence whether it’s worth $2000 or $20? I know I couldn’t.
This is exactly the challenge facing anyone who uses AI to write grant applications without grant assessment expertise. Whilst a person may be able to tell that it reads well, do they know if it actually addresses what assessors need to see, or whether it positions the project in a way that makes funding bodies want to champion it?
Even many professional grant writers face a version of this problem.
They might have spent years working as a writer, mastering how to craft compelling narratives, persuasive writing techniques, and structured arguments, and of course these are all valuable skills. However, without having sat on the other side of the grant table actually assessing the applications, it can be hard for even these experienced professionals to understand the subtleties behind how a successful proposal is picked.
Grant assessment operates according to its own logic. Funding bodies have specific priorities they need to address and assessors look for particular types of evidence presented in particular ways. Assessors also need to justify their scores and funding recommendations to colleagues and stakeholders.
When you’ve been a grants assessor, you gain ‘super powers’ in reading applications. You can see when brilliant projects are described poorly and mediocre projects are positioned brilliantly. You can see the mechanics behind the applications that capture attention whilst hundreds of other similar sounding ones get easily dismissed. You’ve also experienced the heart-breaking frustration of wanting to fund a worthy project but knowing it lacks the evidence in the application to justify awarding a positive funding outcome.
This experience creates something that neither AI nor writing skill alone can replicate – informed judgement and ‘good taste’.
At Whitney Consulting, every grant application undergoes quality assessment by someone who has sat in the assessor’s chair. Tara Whitney, the company’s founder, was a government grants assessor for years before starting the business, and the rest of our senior staff are former government grants assessors too. Between them, they’ve reviewed thousands of funding applications, helping to decide which projects receive money according to specific guidelines and complex scoring systems.
For our clients, this means that before any application goes back to them, it has been evaluated through the lens of assessment experience, not just writing ability. We don’t ask “Does this read well?” We ask “Would this convince an assessor to recommend funding?” That distinction matters profoundly.
So, does this mean AI has no place in grant writing? Not necessarily.
Tools have their uses, and technology continues to evolve. AI can generate text based on patterns it has learned from existing documents and can perfectly mimic structure and style. Critically though, what it can’t do is understand the strategic positioning that transforms a good project into a funded one, nor can it recognise when evidence is technically present but inadequately framed. It cannot anticipate the questions a human assessor will ask when they reach the evaluation stage.
So please take it from us, the people who have the good taste to know what a great application looks like, if you’re considering using AI for your grant application ask yourself how will you evaluate whether the output is genuinely strong, or simply well-written? Who on your team has the experience to make that judgement? If the answer is no one, you’re essentially hoping that an algorithm has correctly intuited the complex, often unstated requirements of grant assessment. That’s a significant gamble when substantial funding hangs in the balance.
We aren’t interested in dismissing technology or claiming that human-only approaches are superior. Instead, we believe that applications are always best when they combine the writing skills of a professional, the subject matter expertise of a client, and the experience that comes from sitting on the assessor’s side of the funding table.
