Three men with big ambitions for building AI factories across Australia have been asking investors for billions. One has a conviction for insider trading. The other is his cousin. A third co-founder is his former brother-in-law. Ninety-seven per cent of what they have promised to build does not yet exist. Hundreds of fund managers packed a Sydney hotel auditorium to hear them. Their planned ASX float was tipped to be Australia’s biggest since Telstra in 1997, and hoped to hit a value just short of $44 billion. But on Friday, they pulled the IPO. It’s incredible how much hype has built around Firmus. Especially what we hear over and over: female founders struggle to get investor attention, and they received just two per cent of Australian venture capital last year, down from four per cent the year before. The hype centred on some ambitious beliefs. Oliver Curtis, his cousin Tim Rosenfield and his former brother-in-law Jonathan Levee founded Firmus in 2019 as a bitcoin mining operation before pivoting to building and operating massive data centres for AI companies. They planned to sell shares to the public for the first time via an ASX float, hoping to raise $7 billion at a valuation just under $44 billion. A year ago, Firmus was valued at $1.85 billion. And just two months ago, it was $15 billion. Analysts have noted that, in the lead-up to the once-tipped late October float, it has around US$30 billion in debt, roughly equal to its aspired valuation of AU$43.7 billion. Firmus has promised 97 per cent of the data centre capacity that hasn’t been built. The company has limited financial history and has provided scant forecasting details. It has never run a data centre of significant scale; it operates two small data centre sites. While it counts Meta and OpenAI as customers, analysts have noted the lack of information on signed contracts and how much these customers actually plan to spend. Jun Bei Liu, portfolio manager at Ten Cap, described it as “one of the most polarising IPOs” in a research note. “The lack of detail they disclose is unprecedented. I think Firmus is a very high-risk proposition—very high profile, but with enormous risk involved.” Liu said her firm prefers to be in companies “that actually build those data centres – the engineers, the electrical contracts…” Still, days before the float was pulled, bankers were reporting substantial interest, according to the Australian Financial Review. Now, compare that with what women founders are asked to prove, and the questions they get well before they’d ever get to anything like this point. As research by Dana Kanze has found, investors tend to ask male founders how they will win and female founders how they will avoid losing. All-women founding teams have secured just three funding deals above the $50 million mark since 2019, according to the State of Australian Startup Funding report. Revenue, contracts and a strong track record are the prices of entry for most women seeking capital. Firmus seems to have skirted the edge of a $44 billion listing by promising to ride the AI boom. The question is who pays if Firmus succeeds in its ambitions. Firmus is currently building a $2 billion data centre near Launceston. It is due to start operating in early 2027 and was recently approved for a second centre at Bell Bay. These factories will make Tasmania’s biggest power user, with a third application lodged for a site in Wesley Vale. 6000 people have signed a petition against the Bell Bay project, raising concerns about noise, energy, water and a lack of consultation. On Thursday, the day before the float was pulled, Firmus withdrew from giving evidence to the federal parliamentary inquiry into AI. I’d argue an “AI Factory” pursuing energy-soaking mega developments across the country should be answerable to such an inquiry, and should do everything it can to secure the social licence it needs. Elsewhere, communities are winning fights against data centre developments – which you’d also think would make investors nervous. Proposed data centres, including GreenSquareDC’s in Hazelmere, Perth, and Goodman Group’s Project Mars (a tip, don’t call such a project “Project Mars) in Sydney’s north, have been scrapped following local opposition. The company’s statement announcing it had pulled the IP blamed market conditions. “Having considered recent market volatility and prevailing market conditions, the board determined that the terms on which the offer could proceed would not appropriately reflect the strength of the company’s business and long-term growth outlook.” It’s now expected to look to the private market for money, including Nvidia and other existing backers. But “market volatility” doesn’t explain much. Instead, a lack of detail does, along with a strong belief that the detail wouldn’t be needed. The parliamentary inquiry should seek to get Firmus back in the room. And the investors who filled that auditorium might want to ask themselves why only two per cent of capital goes to all-female-founded teams, many of which pitch far more detail than Firmus offered, with those women founders asking the same questions. Share this Firmuswomen founderswomen in startups by Angela Priestley 2 hours ago ## Stay Smart! Get Women's Agenda in your inbox * Email This field is for validation purposes and should be left unchanged. * Email SUBSCRIBE ## Latest news ### Matty J on Sam Wood: A good reputation doesn’t make you a good man. ### Billionaire Nicola Forrest donates more than $500 million to her charity focused on women and children ### The women challenging Australia’s mental health system: Meet two recipients of the 2026 Australian Mental Health Prize ### Sam Wood’s career is not the biggest casualty of his alleged violence ### International Day of the Girl 2026: Meet the girls changing the world, from Gaza to Tuvalu ### Podcast: Why it’s time to quit being nice

ALT: Homer Simpson Disappearing Bush Meme