AI x Funding: The Next Wave of Startup Capital

AI, funding, and teams: what founders need to understand right now

AI has shifted the startup landscape faster than almost any technology wave before it. For founders, that shift is showing up in three places at once: how investors assess companies, how teams are built, and how value is created or missed.

What’s changed isn’t just tooling, it’s expectations. Investors are asking different questions. Teams are operating with different constraints. And founders are being pushed to separate real progress from noise.

This article distils the core themes from a recent Cake webinar into what matters most for founders right now — without the hype, and without the play-by-play.

“AI-first” is about tangible value

One of the clearest themes was that “AI-first” has quickly become a meaningless label on its own.

Most companies today can add an AI feature. That’s no longer the differentiator. What investors are looking for is whether AI is meaningfully embedded in the product and business  or whether it’s bolted on without changing outcomes.

In practice, that distinction shows up in a few ways:

• AI that materially improves speed, quality, or cost for the customer
  
  
• Products that rework workflows, rather than just automating a single task
  
  
• Clear understanding of the trade-offs between model performance, reliability, and margin
  

A recurring warning was against superficial AI usage. Adding a chatbot or generative feature without customer pull can introduce real cost and complexity without delivering corresponding value.

The strongest companies treat AI as infrastructure, not a headline. As Yash Varma from Empress Capital put it directly: "AI is not the product." It's table stakes, what matters is what you build around it. The technology should disappear into the experience, while the customer benefit becomes obvious.

Funding looks healthier — but the gap is widening

Recent funding data shows a market that’s steadier than many founders expect. Deal activity has picked up, and valuations across early stages have recovered from their lows.

But that top-line health hides a growing divide.

Companies with credible AI leverage are increasingly seeing valuation premiums, while others are finding fundraising slower and more selective. The market is not uniformly tight, it's uneven.

Asked directly whether current valuations constitute a bubble, Kate Coffey from Airtree offered a candid assessment: "I'm kind of two-thirds it's not a bubble, one-third it's a bubble." The arguments for caution include US capex spending as a percentage of GDP exceeding dot-com era levels. The arguments against focus on the immediacy of efficiency gains - unlike previous hype cycles, AI is delivering measurable value now, not in some projected future.

Another dynamic still working its way through the system is the after-effect of the 2021–2022 funding cycle. Many companies raised at high valuations during that period and have struggled to grow into them. That has made progression into later rounds harder, even for otherwise solid businesses.

Newer cohorts are showing better early discipline. The open question is whether those companies can consistently graduate through Series A and B as expectations reset around growth, efficiency, and defensibility.

For founders, the implication is simple but uncomfortable: fundraising conditions are no longer just about timing the market. They’re about demonstrating real momentum and leverage.

AI is reshaping defensibility

Another theme was how AI is blurring traditional category boundaries.

While much of today’s AI activity sits at the application layer, funding is flowing into areas that combine software with hardware, data, or regulated environments; including robotics, biotech, and climate.

What’s changing is how defensibility is assessed.

Pure software advantages can be replicated quickly when AI accelerates development cycles. In some cases, physical systems, proprietary data, or tightly integrated workflows now create stronger moats than code alone.

That doesn’t mean every company needs hardware. But it does mean founders should think more carefully about what truly compounds over time and what competitors could reproduce faster than expected.

Valuation premiums exist — but aren’t automatic

AI-driven valuation premiums are real, though the clearest data comes from the US market. PitchBook figures show consistent premiums for AI-first companies at each funding stage, with larger step-ups between rounds compared to non-AI peers. Australian data is less granular, but deal sizes (a useful proxy for investor appetite) tell a similar story.

The discussion repeatedly came back to the same drivers:

• AI unlocking markets that were previously hard to serve
  
  
• Products seeing unusually fast adoption because they slot directly into workflows
  
  
• Clear paths to large markets, not just technical novelty
  
  
• Revenue and usage signals that support the story
  
  

There was also a note of caution around anchoring too heavily on mega-rounds or hyperscaler dynamics. Those sit in a different universe from most founder-led startups and aren’t useful benchmarks.

The takeaway for founders is grounded: AI can amplify strong businesses, but it doesn’t rescue weak ones. Performance still does the work.

If you’re not “AI-first”, relevance still matters

Not every company is, or should be, AI-native. But that doesn't mean AI is optional.

One notable shift: AI is expanding the universe of investable businesses. Kate Coffey pointed to services businesses e.g. accounting firms, law practices, that traditionally couldn't achieve venture-scale margins. With the right AI-enabled setup, these businesses can now operate at margins and scale that attract institutional investment. "The investable universe for Airtree is getting bigger," she noted.

For companies without AI at their core, relevance increasingly comes from how AI is used internally and adjacent to the product, rather than from repositioning the business overnight.

The most practical uses discussed included:

• Improving internal efficiency across engineering, sales, recruiting, and content
  
  
• Enhancing existing products where AI meaningfully improves the customer experience
  
  
• Using AI to test, learn, and iterate faster (not to replace strategic thinking)
  
  

Growth remains the primary signal investors reward. AI is one lever among many, but it’s becoming a baseline expectation in how modern companies operate.

Smaller teams, higher output, different metrics

One of the most tangible shifts for founders is how team size relates to progress.

Median team sizes at early stages have fallen, while revenue per employee has risen. AI tooling is enabling individuals to do more (particularly in engineering and product)  which changes how companies scale.

This has knock-on effects:

• Headcount growth is no longer a reliable signal of momentum
  
  
• Lean teams can ship faster than much larger ones could a few years ago
  
  
• Investors are paying closer attention to ARR per employee, not just ARR
  
  

The idea of a one-person billion-dollar company was treated with scepticism. Building enduring businesses still involves partnerships, customers, compliance, and human judgement. But the direction is clear: efficiency matters more than optics.

For founders, this raises the bar on hiring. When teams are smaller, every hire carries more weight - culturally and operationally.

AI talent is scarce and equity matters

The competition for experienced AI and ML talent is intense, and compensation reflects that reality.

Salary premiums are common. Equity expectations are rising too, particularly among candidates who understand their leverage and want ownership, not just pay.

A few practical implications emerged:

• Equity structures don’t need to be one-size-fits-all
  
  
• Vesting, milestones, and role design can be adjusted thoughtfully
  
  
• Culture, mission, and trust increasingly matter alongside compensation
  
  

There was also a clear signal that equity literacy in Australia is improving. Candidates are asking better questions, and founders are being pushed to explain value more clearly so that the team actually understands it. 

Used well, equity remains one of the most powerful tools founders have to align incentives in a competitive talent market.

What founders should take from this

Across funding, teams, and technology, a consistent message came through: clarity beats hype.

AI is reshaping the startup landscape, but it hasn’t replaced fundamentals. Companies still win by solving real problems, building things customers want, and assembling teams that can execute with focus.

For founders navigating this moment:

• Treat AI as a capability
  
  
• Be honest about where it creates value
  
  
• Optimise for learning and efficiency
  
  
• Hire deliberately, knowing each role matters more in smaller teams
  
  

Over time, “AI companies” may stop being a category at all. It will simply be part of how modern businesses are built.

Until then, the advantage belongs to founders who stay grounded and make decisions based on substance, not noise.

‍