The Australian Government has released a formal set of expectations for data centres and AI infrastructure developers, establishing a "social licence to operate" as the foundation for future investments. Aimed at capturing the opportunities of the National AI Plan, the framework priorities proposals that align with Australia’s national interests, security, and data sovereignty.
The expectations mandate that new or expanded developments — including hyperscale operations and "AI factories" — must contribute positively to the energy transition, minimise water usage through circular solutions, and invest in a skilled domestic workforce.
Crucially, the government will fast-track regulatory assessments for proposals that meet these criteria, while energy-intensive projects that do not align with national sustainability and security goals will not be prioritized.
The Five National Expectations for AI Infrastructure
National Interest & Sovereignty: Operators must protect sensitive data, prepare for physical and digital threats, and ensure their business benefits the local economy and community.
Energy Transition Support: Data centres are expected to secure additional clean energy or storage to offset demand, avoid upward pressure on consumer prices, and enhance grid stability through demand flexibility.
Sustainable Water Usage: Encourages innovative cooling technologies and the use of non-potable water, requiring transparent reporting on water efficiency and resilience to climate-driven disruptions.
Skills & Workforce Investment: Developers must create secure, well-paid jobs and establish structured training pathways, including apprenticeships, in collaboration with unions and education providers.
Research & Local Capability: Providers of large-scale compute are expected to enable access for start-ups and researchers on favorable terms, while deploying engineers locally to build Australian supply chain capability.
What is an "AI Factory"? An AI Factory refers to large-scale, high-performance compute centres specifically designed to train and run complex Artificial Intelligence models. Unlike traditional data centres that primarily store and process everyday digital services, AI factories are highly energy and water-intensive due to the specialized hardware required for massive parallel processing. It plays a role in modern industrial strategy by serving as the "engine room" for innovation, scientific research, and national security applications. This concept is supported by the goal of localizing technical expertise, ensuring that the economic and technological benefits of AI are not just imported, but generated within a country's own sovereign digital borders.
Policy Relevance: Insights for India’s Data Centre Revolution
Scaling Green Data Infrastructure: As India projects a 30% surge in electricity demand driven by AI data centres, adopting the Australian model of "demand flexibility" could reflect growth in the stability of India's 520 GW grid.
Internalising Water-Efficient Cooling: With Indian catchments under pressure, the Australian expectation for non-potable water use in data centres plays a role in aligning the Jal Jeevan Mission 2.0 sustainability goals with the high-tech sector.
Bypassing the Automation-Upskilling Gap: The mandate for structured training pathways and local engineering deployments is supported by the need to address the "upskilling gap" identified in UNIDO employment trends for Southern Asia.
Supporting "Aatmanirbhar" AI Factories: Encouraging providers to enable favorable compute access for start-ups contributes to the success of the BHARAT-VISTAAR platform and other sovereign AI initiatives.
Leveraging "Social Licence" for Parks: Applying a similar "Social Licence" framework to the 100 new BHAVYA industrial parks could ensure that large-scale infrastructure projects maintain long-term community buy-in and data sovereignty.
Relevant Question from Policy Stakeholders: How can MeitY (Ministry of Electronics and IT) utilise the "favourable terms for start-ups" model to mechanically increase the number of Indian AI researchers working on indigenous LLMs like BharatGen?
Follow the Full Expectations Here: Australian Government: Expectations of Data Centres and AI Infrastructure Developers - March 2026
