Wilson Asset Management cornerstones SCX.ai’s $40 million IPO with support from Ellerston Capital and other institutionsStrong institutional demand has seen the offer substantially covered ahead of Wednesday’s closeSCX.ai will become Australia’s first listed pure-play sovereign AI inference-as-a-service company
Special Report**: Institutional investors have thrown their support behind Australian AI infrastructure company SCX.ai, with The Australian Financial Review reporting Wilson Asset Management (WAM) is cornerstoning the $40 million IPO.**
The AFR reports WAM will join Ellerston Capital and a number of other institutional investors on the register as the company prepares to list on the ASX under the code SCX.
This offer, jointly led and fully underwritten by Henslow and Canaccord Genuity, has attracted strong institutional demand ahead of its scheduled close on Wednesday evening.
The cornerstone investment comes as investors increasingly look beyond companies building frontier AI models and towards the infrastructure that allows enterprises to run those models securely, efficiently and without vendor lock-in.
The AFR recently reported WAM Active, the fund managed by Shaun Weick, delivered a record 75.5% return in FY26, driven in part by its successful investment in AI infrastructure company Firmus.
Weick told the publication he had positioned the portfolio to benefit from “every stage of the global build-out of AI and its supply chain”.
Wilson’s backing is expected to draw attention to what will become Australia’s first listed pure-play sovereign AI inference-as-a-service company.
Built around AI inference
SCX.ai has built its platform around AI inference – the computing that powers chatbots, AI agents, document analysis and other AI applications used by businesses and governments every day.
Rather than constructing large, purpose-built AI campuses, the company deploys specialised AI processors inside existing Australian commercial data centres.
Chief executive David Keane said investor interest was increasingly shifting towards the infrastructure required to operate AI applications, rather than simply train larger models.
“Training frontier AI models is only one part of the AI economy,” Keane said.
“The AI most organisations use every day is inference. Every chatbot conversation, every AI agent, every document summarised and every workflow powered by artificial intelligence relies on inference.”
The company believes that distinction is becoming increasingly important as debate grows around the environmental footprint of AI infrastructure.
Recent analysis by infrastructure advisory firm Ophir highlights why the debate has intensified. The firm’s modelling estimates Australia’s data centres currently consume about 1.3 billion litres of water a year, but projects that figure could rise to between 4.4bn and 5.9bn litres annually by 2031 if proposed developments proceed.
At the midpoint of those forecasts, data centres would consume more water than power station cooling and golf courses, prompting growing scrutiny of how future AI infrastructure is designed and where it is built.
Need for right type of infrastructure
Keane said Australia should focus on building the right type of AI infrastructure.
“People often assume all AI infrastructure looks the same, but it doesn’t,” he said.
“The AI applications businesses and governments use every day can be delivered much more efficiently using purpose-built inference infrastructure. By deploying that infrastructure inside existing commercial data centres, we can expand Australia’s sovereign AI capability without building a new generation of massive AI campuses.”
The investment thesis has also been reinforced globally as attention shifts from building ever-larger AI models to the infrastructure needed to run them.
Earlier this year, AI infrastructure company SambaNova – whose technology underpins SCX.ai’s platform – raised US$1bn at a US$11bn valuation, reflecting growing investor confidence in AI inference infrastructure.
At the same time, major AI companies including OpenAI and Google have backed the push towards open-weight AI models, allowing enterprises greater flexibility over how and where they deploy AI.
Keane said that trend made sovereign inference infrastructure increasingly important because organisations still needed secure, local compute to run those models.
“Open-weight models are giving organisations more choice than ever before,” he said.
“Every organisation still needs secure, sovereign infrastructure to run those models. That’s where we see the biggest opportunity as AI moves from experimentation into everyday business.”
According to its prospectus, SCX.ai has contracted annual recurring revenue of $5.4m, unbilled total contract value of $14.6m and 298 active users.
Funds raised through the IPO will be used to expand AI infrastructure, accelerate customer acquisition and continue development of the platform.
This article was developed in collaboration with SCX.ai, a Stockhead advertiser at the time of publishing.
This article does not constitute financial product advice. You should consider obtaining independent advice before making any financial decisions.
Investors should only rely on information contained in the company’s Prospectus when considering any investment.
Related Topics #
UNLOCK INSIGHTS
Discover the untold stories of emerging ASX stocks.
Daily news and expert analysis, it's free to subscribe.
By proceeding, you confirm you understand that we handle personal information in accordance with our Privacy Policy.