AI took half of global venture capital in 2025. What it means for everyone else
Artificial intelligence companies raised around half of all venture funding worldwide in 2025, and around 61 per cent in Australia. We look at what the AI wave means for founders outside it, and for those building inside it.
Key points
- AI companies raised about US$211 billion in 2025, roughly half of global venture funding.
- In Australia, AI accounted for around 61 per cent of the capital startups raised.
- Founders outside AI face more competition for attention, and need sharper evidence.
- For AI founders, the bar is shifting from the model to the customer problem and the data.
In 2025, artificial intelligence was not just a sector of venture capital. It was close to half of it.
According to Crunchbase, AI companies raised about US$211 billion worldwide, roughly half of the US$425 billion invested in startups. In Australia, the State of Australian Startup Funding report found AI accounted for around 61 per cent of the capital raised.
Why the money moved
Investors are betting that AI will change how almost every kind of software, and many kinds of work, are done. Some of the largest rounds in history went to companies building foundation models, which need enormous amounts of capital for computing power. Below them sits a fast-growing layer of companies applying AI to specific industries and tasks.
If you are not an AI company
For founders outside AI, the shift has real consequences:
- Attention is scarcer. Investors have limited time, and AI has taken a large share of it.
- Evidence matters more. Without a hot sector behind you, revenue, retention and customer proof carry more weight.
- Everyone gets asked the AI question. Expect investors to ask how AI affects your market, your costs and your competitors. Have a clear, honest answer.
If you are building with AI
For AI founders, the opportunity is large, and so is the competition. A few patterns are becoming clear:
- The model is rarely the moat. Many companies can access similar models. Lasting advantage tends to come from proprietary data, deep workflow integration and distribution.
- The customer problem still comes first. “We use AI” is not a value proposition. “We cut the time to process a claim from days to minutes” is.
- Costs need watching. Computing costs can scale with usage. Unit economics should be modelled early.
“We use AI” is not a business. Solving a painful problem, and doing it better because of AI, can be.
Is it a bubble?
Every technology wave attracts more money than it can use well, and some of today’s valuations will not hold. That doesn’t mean the underlying change isn’t real. The internet boom of the late 1990s overshot, and the internet still reshaped the economy. The useful question for founders and investors is not whether AI is overhyped in aggregate, but whether a particular company solves a real problem in a way customers will pay for.
Sources and further reading
- Crunchbase News, Global venture funding data for 2025. news.crunchbase.com
- SmartCompany, Australian startups raised $5.1 billion in 2025. smartcompany.com.au
Source: Cut Through Venture and Folklore Ventures via SmartCompany; Crunchbase News
General information only. This article reflects the author’s views at the time of writing. It does not take into account your objectives, financial situation or needs, and it is not financial, legal or tax advice or a recommendation to invest. Figures are from the sources named and may change. See our terms of use.