From Boom to Bust in Five Days: The Firmus Collapse Explained

By Marcus Bennett

Just over a week ago, the Australian AI company Firmus was preparing to list on the stock market with a valuation that would have placed the startup in the same league as Woolworths, one of Australia's largest retailers. Last Friday, the float collapsed.

The Guardian has examined the collapse in a new podcast episode, "How Firmus went from Australia's AI 'fairytale' to flop," in which business editor Jonathan Barrett speaks with reporter Matilda Boseley about the hubris of the firm's founders and what the failure signals for the wider AI investment bubble.

What happened to the Firmus float?

The trajectory was steep. As The Guardian framed it, the company went from "boom to bust in five days." The listing was billed as an Australian business success story — a local AI firm reaching a market value comparable to Woolworths, a cornerstone of the Australian share market.

Then, on Friday of last week, the deal fell apart.

The Guardian's companion reporting, published October 9, 2026, documented the implosion of the "much-hyped Australian stock market listing." A second report, published October 3, 2026, had already raised questions about the multibillion-dollar float amid a backlash against the datacentres (the physical facilities that house the computing infrastructure AI systems require) underpinning the company's plans in Tasmania.

That earlier piece carried a blunt assessment of the founders' standing: "These guys are just coming from nothing."

Who is Oliver Curtis?

At the centre of the collapse sits Firmus co-founder Oliver Curtis, an entrepreneur attempting to rebuild his professional reputation after serving time in prison for insider trading (the illegal practice of trading on non-public market information).

Curtis has offered his own explanation for the failure. He attributed the collapse to Australia's "mantra of disbelief around entrepreneurs," according to remarks reported by The Guardian.

The Guardian's podcast discussion, led by business editor Jonathan Barrett, probes what the outlet describes as "the hubris of the firm's founders" — a framing that stands in direct tension with Curtis's claim that Australian skepticism, not the founders' conduct, killed the deal.

Why does the collapse matter beyond Australia?

The stakes extend past one company. Firmus was positioned as proof that an Australian AI startup could command a valuation alongside the country's corporate giants. Its collapse removes that proof point.

The Guardian connects the failure to two larger pressures.

First, the AI investment bubble. Barrett and Boseley's discussion treats Firmus as a case study in what happens when hype outpaces substance — a multibillion-dollar valuation attached to a company whose float drew the descriptor "coming from nothing" from critics quoted in the outlet's October 3 report.

Second, the datacentre backlash. Firmus's plans in Tasmania faced public resistance, and The Guardian's reporting situates the failed float within that broader fight over the physical footprint of AI infrastructure. The outcome suggests that datacentre opposition is now a material risk for AI companies seeking public market funding, not merely a planning nuisance.

What questions remain open?

The podcast does not close the file on Firmus. The Guardian's framing — a "fairytale" turned "flop" — leaves the central dispute unresolved: whether Australia's investment culture failed an ambitious founder, or whether investors correctly priced the risks in a company led by a formerly imprisoned executive and pursued a multibillion-dollar listing amid datacentre opposition.

Curtis's own words anchor one side of that argument: the "mantra of disbelief around entrepreneurs." The Guardian's reporting on founder hubris anchors the other.

For the AI sector, the significance is straightforward. A company once valued in the same league as Woolworths lost its listing in five days. If similar floats falter elsewhere, the Firmus case will serve as the reference point for how quickly AI valuations can unwind when public-market scrutiny replaces private-market enthusiasm.