Your pitch deck doesn’t need more AI. It needs more of you.

There’s a big problem in the way founders are using AI for fundraising.

Sure, the tools are getting better. They’re faster, more capable and able to produce something that seems highly polished in seconds. Yet the more founders use AI to build their decks, the more every deck starts to sound the same.

Yes, the language is cleaner, the formatting is neater and the phrasing sounds more “investor-ready”, but somehow the result is far less persuasive.

That’s because in a world of automation, cut-through now comes from humanisation.

If you’re raising seed investment this really matters, probably more than you realise. Investors aren’t backing a sequence of well-constructed prompts. They’re backing a person. More specifically, they’re backing you. They’re backing your judgement, your clarity, your instinct and your ability to roll with the punches.

AI can help you do the mundane work. It can speed up research, process information and reduce hours of manual effort. I use it heavily myself. I have trained custom GPTs to support parts of my work because it’s a powerful tool. But there’s a line. And that line is being crossed by too many founders. They’re not using AI for leverage, but to replace their critical thinking. And that’s where the problem starts.

The autopilot mistake

The way we use AI is much like flying a plane on autopilot.

Autopilot, when launched, was a transformational technology. It handled all the routine parts of a flight, and it can make the process of flying safer and more efficient. But you still need a human for the most valuable parts – the take-off and the landing. Those are the moments that matter most. Those are the moments when judgement and experience come into play.

Fundraising works the same way.

If you run your entire fundraising process on autopilot, using AI to generate the story, shape the strategy and write the words, you remove the very thing investors are trying to assess. You. They want to know whether you can think clearly, whether you understand your market properly and whether your strategy is real or just well-presented.

They’re investing in the pilot, not the automation.

Too many founders are trying to use AI to cover up their gaps. They’re unsure what investors want to see or are not confident in how to communicate their business. So they ask ChatGPT or Claude to do the thinking for them. And the output always looks great, and it usually sounds credible at first glance. And that is the massive trap. It sounds so plausible that founders bank on it, and that works right up until someone asks a follow-up question.

Then it all comes unstuck.

Why this becomes obvious so quickly

A lot of the founders I work with are building AI products themselves, so it’s completely natural that they turn to AI to help build out their investment memo, financials and pitch materials. It’s usually their first instinct.

But as soon as I ask a question to explore deeper, I often get an AI-generated answer back. Straight away, you can tell they’ve not really thought it through. Instead, they’ve thought, “I’m not sure, I’ll get AI answer that for me.”

So as soon as I ask these sorts of questions in a live conversation, they get found out, they can’t hide behind the output anymore. As a result, they fail to articulate what they meant. They can’t explain the reasoning behind the information they shared, and they can’t defend their assumptions. What looked polished in the deck falls apart because there’s no depth behind it.

That’s what over-leveraging AI does. It erodes critical thinking. Founders all become a homogenous group sharing the same tactics and strategies articulated in the same robotic tone.

The trouble is, AI is trained on what is broadly available. It triangulates all the data points to form an “opinion” on what is right. And this includes an enormous amount of advice, decks, articles and templates. The same advice that the 99% of founders who fail to raise funding have been consuming via blog posts and accelerator programmes for the past decade or more.

And this really matters because the goal of pitching is not to make you look like the average founder – the 99% who fail. The goal is to make you look like the top 1% who succeed.

When AI fails to distinguish between the mass advice that’s been guiding founders to raise unsuccessfully for years, and the tiny amount of advice that genuinely helps founders raise, you end up in a position where founders who should be positioning themselves as a 1 in a 1,000 generational talent are putting blind faith in a robot who relies on the law of averages.

When you use it carelessly, AI just pulls you towards the middle. In fundraising, the middle is where most visions die.

If AI was the answer, fundraising success would be rising

This is a question I think more founders should consider…

If AI was genuinely better than a human at helping you connect with an investor and persuade them to part with their capital, why are founders still struggling so much to raise?

Why are we not seeing a dramatic increase in success rates now that so many people have access to these tools?

The fact is, we’re not seeing that.

You can argue, rightly, that market conditions are tougher now than they were five years ago. That is true. But if AI was fundamentally changing the quality of founder communication, there should still be some obvious lift in success rates.

But there isn’t, because founders aren’t leveraging it in the way they should be.

It’s making decks more uniform. It’s reproducing the same stale thinking and it’s turning a founder’s passion into something corporate, robotic and less memorable.

For me, this is exactly why humanisation is now the advantage.

How to use AI without losing your edge

I am not arguing that you should avoid AI. I’m arguing that you should use it properly.

Of course, you should use it in preparing your round where it genuinely creates leverage. Things like processing customer interview transcripts, spotting patterns in data, and uncovering hard-to-find desk research are all great use cases. This cuts manual work that would otherwise take hours or weeks. In some cases, it identifies patterns you’d never had spotted. This is where AI is at its strongest. And this can give you incredible insights you can use to build your credibility with investors and give you a real edge.

But when it comes to your deck, the right way to use AI is to start with depth, not shortcuts.

When creating content, us humans rarely achieve the best articulation of a concept immediately. Ask any advertising executive. They’ll tell you the brochure is much easier to write than the billboard. Distilling the message down to its core takes a process of distilling information.

You need to guide any AI you use through that process. Doing so allows the AI to act more closely to the human mind.

Step 1: Do the hard yards

Before you even think about slide design or tidy bullet points, map your pitch like a business plan.

For each slide, write at least 500 words explaining what is strategically important. Do the messy thinking yourself. Explain the problem fully. Explain the solution fully. Write out your reasoning, your traction, your roadmap, your go-to-market approach.

This is the work most founders try to skip, and it’s the work that creates the substance investors are actually looking for.

Step 2: Use AI to distil the core points

Once you’ve laid down your strategic thinking, use AI to distil the information (I find Chat GPT is best for this). Feed in those long-form sections and ask it to pull out the most relevant points as single-line bullets using the below prompt. That gives you a launch pad for slide content.

You are a pre-seed / seed investment consultant with deep experience in the [country] funding landscape. I will shortly upload notes about key sections of my business model that are most relevant to an investor pitch, which you will use as the basis of the output. With the notes provided, produce a concise bullet point list distilling the content into a list of 6-12 word bullet points. You should consider the key elements that are most relevant and impactful with investors. Reduce the content down to a maximum of 10 bullet points. Keep it plain English. Reflect the inputs faithfully, without contradicting or replacing their core concept. Do not invent new information; keep strictly to the information provided.

This way, you’re going from your deep strategic thinking to condensed clarity. Rather than surface-level thoughts to polished garbage.

Step 3: Write your deck

Use the distilled points to create the first draft of your deck by yourself. Use the most compelling and impactful point as your headline. This hooks the reader. Select other points to form the core content of your pitch deck. But be selective. Your goal is to engage, not educate. So, only select the information that you believe will lead the investor to want to take the conversation further with you.

Step 4: Refine with AI
Take your deck back into an AI tool (I find Claude is best for this). Start with the following prompt, then feed in the draft content of your deck.

You're my copywriting assistant. I'm going to share with you some draft slide content for an investor pitch deck. I need you to enhance the content by making it shorter, punchier and more memorable where possible. The tone should be bold, disruptive and credible. All slides need to elevate the value proposition and make the product and opportunity sound exciting to Angels and VCs. We're aiming for slides of 50-75 words a slide. Please do not repeat the same headline structure too often. The output should match the text structure of the input. For example, if the input is bullet points, the output should also be bullet points. Avoid headlines that use the rule of three. Ensure all headlines act as a summary of the slide, allowing investors to quickly understand the key takeaway for each slide.

The output will not have nailed it. The correct slide is probably somewhere between the Claude output and your original. Strategically select the improvements and keep the content you wrote that you think is better.

WARNING: When following this process, you’ll need to be wary of the language trap. AI loves financial buzzwords. It writes as though every investor is an institutional VP reading a board memo. But most pre-seed and seed investors are founders and entrepreneurs themselves. In many cases, the best way to explain your business is the way you would explain it to a mate down the pub.

The real problem for founders

Founders are under pressure. There is so much to do and not enough time to do it. You’re expected to move quickly, cover every role and somehow still show up as strategic, thoughtful and investor-ready. So, of course, you look for shortcuts.

You should be looking for shortcuts in manual processes, ways to streamline anything repetitive and systems that can automate junior tasks.

But fundraising is not a junior task.

At its core, fundraising is human. It’s one person trying to get another person excited enough to believe and trust in their idea. That means your job isn’t just to assemble information, it’s to enrol investors into your vision. Only a human can do that properly.

A lot of founders use AI because, underneath it all, they’re simply unsure. They don’t fully know what investors are looking for. They don’t trust their own instinct enough. And with all of the world’s knowledge available at the click of a button and presented as customised to you and your business, it’s clear to see why you might think AI will somehow do a better job of sounding investable.

But more often than not, it just strips out the one thing that would have made your pitch compelling in the first place – you.

That’s why I think more founders need to break free of the idea that AI can improve the essence of their pitch. It can improve the processing and the efficiency. But it can’t improve the human core unless that core is already there.

What I would do next if I were in your position

Take your current deck slide by slide and force yourself back into the substance.

Follow the process outlined above. Do the hard yards and then use AI to help you boil it down. Write the pitch deck yourself without AI assistance and then leverage AI to sharpen where needed.

That process is slower than prompting a tool to write ten slides for you in 5 seconds. But it’s also far more likely to produce something worth backing.

The founders who stand out now aren’t the ones using the most AI. They’re the ones using it with restraint. They let it handle the automation while they hold on to the human connection. And in pre-seed and seed stage fundraising, the human connection is the thing that matters most.