Grant, tender and award submissions: to AI or not to AI?

TL;DR: In my experience, AI isn’t better at this yet and the risks of using AI for confidential or sensitive submissions may outweigh the perceived benefits. If you want the strongest possible submission, you’ll need an experienced, strategic, trusted human, if you have the option.

It’s definitely the question of 2026: is it okay to use a generative AI tool to write your grant, tender or award responses?

It was the industry question in January when I made a return to focusing on submissions writing, after 5 years of broader communications work in both the community services sector and the private sector.

It’s still the question in July, as I reflect on the past 6 months.

Anyone engaging with the question might be asking not only if it’s efficient but if it’s allowed, encouraged, expected or ethical.

The sharpest questions for me are: Is it competitive? What are the risks? Does it align with organisational values, strategies, goals and executive direction?

Excellence

Honestly, in January I wasn’t sure anyone would still pay me to do what I do best (which is to secure your next organisational win with strategic submissions, compelling narratives and related communications).

Fortunately, for now at least, this human is still able to invoice from the ikigai intersection of what I love, what I’m good at, what the world needs, and what I can be paid for.

That intersection, the space I work in and what all of my clients have in common, is business or organisational excellence. That’s the whole point when it comes to a bid, a grant application or an award entry.

So if we’re doing excellence, our tools and other choices will either support or detract from the mission.

This year, where client permissions permitted, I tested one AI tool's ability to improve on my strongest criterion responses and overall submissions.

Every time, the AI response was a dull, diluted version of the story, sometimes losing crucial data or very intentional wording, designed to connect with the reader on a human level. I also noticed other issues where it just wasn’t up to the job.

“Every time, the AI response was a dull, diluted version of the story”

Excellence can also relate to environmental impact, data security, social responsibility, community impact and more. That said, it’s getting difficult to avoid AI - it’s in my search engine, social media, design tools - and I believe the heaviest burden of responsibility should sit with the giants who are the primary consumers and drivers of AI.

Emotion

Human-to-human has always been my approach, and for now at least, assessors and judges are still human. This has always been one of my trade secrets - I write for the audience, even if it’s an audience of one. I write for the human who will be reading. I take time and care to understand funder or panel priorities but also to ensure an engaging read - the true need or impact, the most compelling aspects of the story, artful word choices, and no unnecessary words.

Most importantly, the reader should feel something - moved enough to want to retell your story. I prefer to go a step further and ensure they immediately have the words they can imagine using on stage or in their morning meeting, as they introduce the winner to the audience or shortlist to the team.

Chances are your assessor has a dozen submissions to read by 5pm, needs to pick up the kids and dinner on the way home and has endured their least favourite buzz words (or ChatGPT’s favourite expressions) just as many times (or a variation of that kind of day!). Let your submission be a relief, and carry the labour a little.

Also, until assessment teams are relying on AI to select successful applicants, think of your submission as an extension of your brand identity. My approach has always been brand-led or customer experience-led. Your submission is another touchpoint for a strategically cultivated, customer-centred experience. Consider what damage can be done to your brand (trustworthiness, confidence, integrity, authority) when your submission reads like every other AI-generated entry?

Exceptions

It’s not that you can’t have a successful outcome with AI-written responses, you simply risk submitting a less competitive, more generic submission.

Sometimes you have to do the best you can with what you’ve got. If you don’t have the resources, whether that’s people, capacity, time, or money to outsource, and you’ve got an opportunity with a deadline, your best option may be your only option.

If everyone’s doing it, then the playing field is level, as they say. However the game changes with the skill of the person feeding the raw material, data and prompts to the machine, and their ability to appraise it and refine it.

It comes back to risk and competitiveness. If you would ideally have your best person writing your submissions, for your best shot, have your best, strategic and risk-aware AI wrangler on the task if using AI.

And hey, no reason why you can’t use AI for brainstorming, summarising, translation, plain English, accessibility, workflows, checklists and other help, if it’s secure and aligned with your organisation's values and direction.

Evolution

I believe I have an uncomfortable responsibility to stay informed and capable with regard to developments in generative AI, including its benefits, drawbacks and broader social and environmental impacts.

For now, I’ll be staying in touch with assessors and award organisers to find out what kinds of submissions they’re seeing in 2026 and what they want to see less of, and more of. Are AI tools improving submissions or simply making them sound more alike? How are assessor panels utilising AI, and where is it heading?

Let’s see where we are in December.