Three months ago, a founder I mentor walked into a pitch meeting with Blackbird. She’d rebuilt her entire deck the night before. Ripped out every mention of “AI-powered” and “machine learning.” Replaced them with customer outcomes, dollar figures, and time saved.

She closed $3.2M in two weeks.

The same deck — with AI branding front and centre — had been rejected by eleven investors over four months. Same product. Same traction. Same founder. Different framing.

That’s not a coincidence. That’s a market signal, and most founders are missing it.

The AI Fatigue Is Real

I’ve been on both sides of the table for two exits, and I’ve never seen investor sentiment shift this fast. In 2024, slapping “AI” on your pitch deck was basically free money. In early 2025, it was still a tailwind. By mid-2025, it was neutral. Now? It’s actively working against you.

Here’s what’s happening: VCs and angel groups have funded hundreds of AI startups over the past two years. The early cohorts are coming due. And the results aren’t pretty. Most are struggling with retention, burning cash on API costs, and failing to demonstrate any real moat.

TechCrunch reported that AI startup failure rates in 2025 outpaced the broader startup market by nearly 2x. Investors aren’t stupid. They see the pattern.

So when you walk into a meeting and your first slide says “AI-Powered Platform for X” — you’re not exciting anyone. You’re triggering pattern recognition for a category that’s disappointed them repeatedly.

What Smart Founders Are Doing Instead

The founders getting funded right now are doing something counterintuitive: they’re burying the AI.

Not removing it. Not pretending it doesn’t exist. Just leading with what actually matters — the problem and the outcome.

One of my portfolio companies helps commercial property managers reduce energy costs. Their product runs on a pretty sophisticated ML pipeline that analyses building usage patterns and adjusts HVAC systems in real time. The tech is genuinely impressive.

But their pitch doesn’t mention any of that until slide seven.

Slide one: “Commercial buildings waste $14B in energy annually in Australia.” Slide two: “Our customers cut energy costs by 31% in the first 90 days.” Slide three: customer testimonials with specific dollar amounts saved.

By the time they get to the AI explanation, investors are already leaning forward. The AI becomes the “how” — not the “what.”

That’s the difference. Nobody funds a technology. They fund a business that happens to use technology.

The “AI Company” Label Is Actually Hurting You

When you brand yourself as an AI company, three bad things happen.

First, you invite the wrong questions. Instead of asking about your market and unit economics, investors grill you about model risk, API dependency, and competitive moats. Valid questions, but they shouldn’t dominate a seed-stage conversation.

Second, you get compared to everyone. There are roughly 4,000 startups in Australia with “AI” in their pitch materials right now. You’re not standing out. You’re dissolving into a crowd.

Third, you anchor on the wrong value. When the underlying models get commoditised further, your positioning evaporates. You built your brand on the ingredient, not the meal.

Real Talk From the Other Side

A mate of mine works with AI consultants Sydney and they said the same thing — clients don’t care about the model, they care about the result. Nobody walks into a meeting and says “I need a transformer-based solution.” They say “I need to process invoices faster” or “I need to reduce customer churn by 20%.”

Investors think the same way. They’re buying an outcome. The AI is a means to that outcome, and it should be presented as such.

How to Restructure Your Pitch

If you’re sitting on a deck that leads with AI, here’s what I’d do this week:

Kill your first three slides and rebuild them. Start with the problem. Use dollar amounts. “Australian manufacturers lose $2.3B annually to unplanned equipment downtime” is infinitely better than “AI-Powered Predictive Maintenance Platform.”

Move all technical architecture to an appendix. Have it ready for due diligence. Don’t lead with it.

Replace “AI-powered” with the specific outcome. Not “AI-powered customer insights” but “reduces churn by 24% in 60 days.” Let the result do the talking.

According to Startup Daily, Australian startups raised record capital in 2025 despite the global slowdown. The money’s still there. But investors are getting smarter about where they put it.

The Bottom Line

I’m not saying AI doesn’t matter. It obviously does. I’ve got seven AI-heavy companies in my current portfolio, and three of them are my best performers.

But none of them pitch themselves as “AI companies.” They pitch themselves as companies that solve expensive problems. The AI gets mentioned in passing, like it’s table stakes. Which it increasingly is.

If your entire identity is “we use AI,” you’re one OpenAI product update away from an existential crisis. If your identity is “we save manufacturers $400K per year in downtime costs,” it doesn’t matter what’s under the hood. The value stands on its own.

Stop selling the engine. Start selling the destination.