Public consultation insights on Getting it right: Building AI infrastructure that works for Australia

AI in Australia's interests

Artificial intelligence (AI) is transforming the way Australians work, learn, create and innovate. It brings new opportunities for economic growth, productivity, scientific discovery and better services. We want to embrace the opportunities of AI but make sure it is used safely and responsibly.

To achieve this, the Australian Government is developing Australia’s AI standards. The standards will be a national framework for the development of AI models and the infrastructure that enables it.

To support this work, the government undertook a public consultation to inform the development of the standards. The consultation, Getting it right: Building AI infrastructure that works for Australia, sought views on how Australia can support AI investment, strengthen national capability and maximise community and economic benefits.

Consultation findings will be considered alongside ongoing engagement, technical evidence and cross-government policy development to inform the design of Australia's AI standards. The Office of AI will continue working with departments, jurisdictions and stakeholders to refine the standards and support a framework that is effective, practical and aligned with the Australian Government's AI objectives.

Process for developing Australia’s AI standards

Consultation

Public consultation conducted
Feedback received from 1,533 stakeholders

Evidence base

Consultation findings
Technical analysis
Targeted consultation

Standards development

Development of proposed requirements and implementation options.

Government consideration

Cross-government policy development.

Introduction of standards

Australia's AI standards to be released.

Who we heard from

The consultation received a total of 1,533 submissions. The largest share of submissions came from:

Individuals (73%)
Business and industry representatives (12%)
Organisations and institutions (8%)*
*includes community groups, not-for-profits, research and education institutions.
Al and technology representatives (6%)

Of submissions with an identified location, the largest shares came from:

New South Wales (225)
Victoria (194)
Queensland (120)

International submissions accounted for approximately 2% of identified submissions, predominantly from the United States and United Kingdom.

What we heard

Common insights across submissions

Standards and safeguards

Stakeholders support nationally consistent standards and generally favour government action in this area

Public trust

Public trust is viewed as essential to successful AI development and deployment.

Environmental impact

Environmental sustainability, particularly energy and water impacts.

Transparency and accountability

Transparency, accountability and enforcement of standards are viewed as equally as important to the standards themselves.

Tangible benefits

Stakeholders expect AI development to deliver tangible benefits for Australians while appropriately managing risks.

Support for nationally consistent AI standards

Stakeholders recognised the significant economic, productivity and innovation opportunities associated with AI, while emphasising that future AI development should occur within a framework that protects communities, the environment and the public interest.

Most stakeholders supported a national framework and saw government action as necessary for clarity, consistency and public confidence.
Industry stakeholders emphasised the importance of clear, predictable, technology-neutral requirements that support investment and maintain international competitiveness.
Community and environmental groups, and individuals argued for strong safeguards to manage environmental, social and safety risks.

Support for considered thresholds

Stakeholders consistently indicated that standards should be applied based on impact of infrastructure and AI activities, rather than a single threshold measure.

Common factors for consideration included:

electricity demand and consumption
water use
compute capacity
physical facility scale
environmental impacts
community impacts
cumulative impacts across facilities and projects.

Many submissions also raised concerns that narrow thresholds could see project fragmentation and allow significant developments to avoid coverage.

Expectations of transparency, accountability and benefits

Stakeholders held strong expectations for standards supported by:

public reporting and disclosure
independent assurance and oversight
clear compliance and enforcement mechanisms
ongoing review and improvement processes.

Communities consistently highlighted the importance of:

Early engagement before decisions are made
Ongoing consultation while the facility is operational
Clear information about impacts and benefits
Tangible local benefits from AI infrastructure investments

Energy, water and local impacts emerged as the most significant issues raised regarding large-scale AI infrastructure and data centre development. There was broad support for infrastructure growth where it is environmentally sustainable, uplifts Australian capability and delivers benefits for communities.

Location and community impacts

Stakeholders identified location as a key determinant of environmental, social and infrastructure impacts. Feedback placed importance on:

aligning infrastructure development with local energy and water availability
understanding regional environmental conditions
engaging communities early and local participation
considering cumulative impacts where multiple facilities are within a region
ensuring communities early share in the benefits of infrastructure investment, including employment opportunities, workforce and skills development.

Energy

Energy impacts were one of the most frequently discussed issues across all stakeholders – addressed in 78% of all submissions. Common themes included:

AI infrastructure should support Australia's energy transition
new facilities should contribute to additional renewable energy capacity
operators should support storage, firming and grid reliability
growing AI-related electricity demand should not increase costs for households or existing businesses
infrastructure should be located where energy systems can accommodate additional demand.

While stakeholders broadly agreed on these objectives, views differed regarding whether requirements should be mandatory or outcome-based.

Skills and training

Stakeholders identified opportunities for data centre operators to strengthen Australia's workforce by:

expanding Australian skills and capability
creating opportunities for under-represented groups
partnering with TAFE, universities and other training providers.

Industry stakeholders favoured flexible outcome plans, while non-industry groups often argued for enforceable targets, public reporting and commitments for operators to fund training capacity.

Water

Water management emerged as a critical concern (addressed in 77% of all submissions), particularly for regional communities and environmental stakeholders. Common themes included:

transparent reporting of water use
use of recycled, stormwater and non-potable water sources
limited reliance on drinking water supplies
consideration of local water conditions and environmental pressures
cumulative impacts in water-stressed regions

Industry stakeholders generally supported water reporting obligations while emphasising flexibility to adopt emerging cooling technologies and innovative approaches.

Stakeholders strongly support requirements that promote safe and secure AI development, while ensuring Australia develops the skills, research capability and governance settings needed to realise the benefits of AI. Views on copyright and training data remain areas that diverge most significantly.

Skills, training and capability

Stakeholders highlighted the need for AI development to contribute to Australia's long-term capability and workforce development. Priorities included:

skills and workforce development
research capability, education opportunities and support for public interest research
greater access to computing resources
building sovereign capability and technical expertise
ensuring Australians share in the benefits created by AI investment.

Universities and research organisations frequently noted that access to advanced compute remains a national constraint and argued that investment should help grow Australia's research and innovation ecosystem.

AI safety and security

There was strong support for mandatory safety and security requirements applying to advanced AI development activities.

Common proposals included:

risk management frameworks
cybersecurity requirements
security requirements for advanced AI systems
incident reporting obligations
independent testing and evaluation
human oversight and accountability mechanisms.

Many submissions urged standards should remain adaptable to respond to future technology developments.

Copyright and AI training data

Creative sector stakeholders commonly argued that:

creators should have visibility over how their works are used
consent mechanisms are necessary
rights holders should be appropriately remunerated for material used in training.

Technology companies and businesses commonly emphasised:

the importance of regulatory certainty
practical and scalable compliance arrangements
maintaining international competitiveness
avoiding requirements that unnecessarily discourage innovation.