TL;DR
- The gist: OpenAI has launched a major Australian expansion, partnering with NEXTDC for infrastructure and local giants for mass workforce upskilling.
- Key details: The deal includes a 550MW hyperscale campus in Sydney and aims to train 1.2 million workers starting in 2026.
- Why it matters: Itsecures sovereign compute capacity outside the US and entrenches OpenAI’s ecosystem into finance and retail sectors.
- Context: The strategy diversifies OpenAI’s supply chain amid rising hardware costs and a projected $207 billion funding gap.
Opening a new front in the global battle for compute capacity, OpenAI has selected Australia for its latest infrastructure expansion. The AI lab signed a memorandum of understanding with NEXTDC, a move that immediately sent the data center operator’s stock surging nearly 11%.
Anchoring the deal is a plan to develop a “hyperscale” AI campus at the S7 site in Eastern Creek, Sydney. Targeting a potential capacity of 550 megawatts (MW), the facility aims to secure sovereign compute resources as part of the broader “OpenAI for Countries” initiative.
Beyond physical hardware, the strategy includes a significant workforce entrenchment play. Partnering with Commonwealth Bank (CommBank), Coles, and Wesfarmers, the company plans to roll out AI training to over 1.2 million Australian workers and small businesses by 2026.
Promo
The Physical Layer: Hyperscale Ambitions in Sydney
Central to this new partnership is the development of the S7 site in Eastern Creek, a strategic location within Sydney’s industrial corridor. OpenAI and NEXTDC have formalized their collaboration through a memorandum of understanding, outlining a roadmap for a facility designed specifically for high-density AI workloads.
Targeting a potential capacity of 550MW, the campus represents a significant escalation in regional infrastructure capabilities. Such scale is necessary to support the training and inference demands of next-generation models, which require uninterrupted power and cooling at levels far exceeding traditional enterprise data centers.
Aligning with the OpenAI for Countries program, the project aims to build sovereign AI capabilities directly within partner nations rather than relying solely on centralized US hubs.
Formalized through a Memorandum of Understanding (MoU), the partnership integrates NEXTDC into the broader “OpenAI for Australia” initiative. The agreement focuses on establishing what the companies describe as “sovereign AI infrastructure”—computing power that is physically located and managed within the country’s borders.
Under the deal, the two entities will collaborate on the full lifecycle of the project, from initial planning and development through to ongoing operations. The focal point of this effort is the construction of a next-generation hyperscale campus at NEXTDC’s S7 site in Eastern Creek, Sydney. This facility is slated to host a large-scale GPU supercluster, providing the specialized hardware necessary to drive advanced model training and inference workloads locally.
Investors reacted violently to the news, driving NEXTDC shares up approximately 11% to reach a peak of A$14.90. The market views the deal as a validation of the company’s aggressive capital expenditure strategy, which has focused on securing land and power ahead of demand.
Validating this sentiment, Michael McCarthy, CEO of AU & NZ at Moomoo, noted that “the guaranteed revenue stream will likely see investors queueing to get involved.”
Structurally, the deal positions OpenAI as an “initial offtaker,” a critical designation that provides a guaranteed revenue floor for the project. This financial certainty allows NEXTDC to accelerate construction timelines and secure necessary funding with lower risk.
The Human Layer: Enterprise Entrenchment
Beyond hardware, OpenAI is executing a significant “soft power” play by embedding itself into the Australian workforce. Partnerships with Commonwealth Bank (CommBank), Coles, and Wesfarmers target non-tech sectors like finance and retail, ensuring deep market penetration.
The initiative aims to upskill 1.2 million workers and small businesses, effectively creating a trained user base for OpenAI products. By integrating its tools into the daily operations of major employers, the company secures long-term dependency and data feedback loops from diverse industries.
To execute this vision, OpenAI has enlisted three of Australia’s most dominant corporate entities: Commonwealth Bank (CommBank), retail giant Coles, and industrial conglomerate Wesfarmers. This coalition forms the backbone of a new skills initiative designed to democratize access to technical literacy.
The program’s ambition is vast, aiming to deliver essential AI training to more than 1.2 million workers and small business owners, effectively ensuring that a significant portion of the national workforce is equipped to utilize the new infrastructure being built.
CommBank’s involvement highlights the specific gap between AI potential and small business adoption. Many smaller enterprises lack the resources to experiment with new technologies, creating a digital divide that this program seeks to bridge.
Addressing this challenge, CommBank CEO Matt Comyn stated that “small businesses are the backbone of Australia’s economy and the engine of our communities, but too many small business owners tell us they simply don’t have the time or confidence to explore how AI could help them.”
Scheduled to begin in 2026, the nationwide rollout of the OpenAI Academy marks a long-term commitment rather than a short-term public relations effort. This strategy mirrors Microsoft’s enterprise playbook, locking in corporate ecosystems early to establish a dominant market position before competitors can gain a foothold.
Strategic Context: The Global Infrastructure War
The deal represents a shift from centralized “superfactories” to distributed, sovereign infrastructure. It contrasts with Microsoft’s approach of building large-scale, continent-spanning clusters in the US, such as a continent-spanning AI superfactory designed to train trillion-parameter models.
The move also echoes SoftBank’s vision, which emphasizes securing power and compute globally. Masayoshi Son recently sold a substantial stake in Nvidia to fund the Stargate infrastructure project, highlighting the intense capital requirements of the current AI landscape.
OpenAI is under immense pressure to secure capacity, facing a projected $207 billion funding gap by 2030 due to spiraling compute costs. Diversifying its supply chain away from total reliance on Microsoft Azure allows the company to mitigate risks associated with centralized dependencies.
Framing the initiative as a catalyst for national development, OpenAI CEO Sam Altman stated that “through OpenAI for Australia, we are focused on accelerating the infrastructure, workforce skills and local ecosystem needed to turn that opportunity into long-term economic growth.”
By partnering directly with regional providers like NEXTDC, OpenAI not only secures necessary compute but also navigates increasingly complex data sovereignty regulations that require sensitive information to remain within national borders.
Market Reality: Boom or Bubble?
While the stock market cheered the deal, analysts warn of potential long-term risks. The “AI Bubble” narrative remains a backdrop, with fears that infrastructure spending is outpacing actual revenue generation across the sector.
Offering a counterpoint to the initial stock surge, McCarthy cautioned that “the deal fits nicely with current market narratives around AI, and NEXTDC shares could spike today, but the glee with which investors may greet this deal could sour over time as reality bites.”
Hardware costs are rising, complicating the economic equation for new facilities. Recent reports indicate tripling memory prices due to OpenAI’s own supply chain demands, which could inflate the capital expenditure required to bring the S7 site online.
Microsoft’s previous a $5 billion investment in Australia serves as a baseline for comparison, showing the scale of capital required to make a meaningful impact in the region. The success of the project hinges on whether the “sovereign AI” demand materializes as predicted or if it remains a theoretical construct driven by speculative investment.
Altman remains optimistic about the local potential, noting that “Australia is well placed to be a global leader in AI, with deep technical talent, strong institutions and a clear ambition to use new technology to lift productivity.”

