Social licence in the age of the AI infrastructure: why building trust matters for AI data centres

The conversation around Australia’s rapidly expanding data centre sector is moving quickly.

In recent weeks, governments have sharpened their focus on one of the industry’s most significant challenges: how to enable the infrastructure needed for Australia’s AI ambitions without shifting the costs onto the communities asked to host it. On 5 August, the Australian Energy Market Commission recommended that data centres bring new clean, firmed energy into the system, operate flexibly and bear the costs and risks created by their connections so existing electricity consumers are not worse off.

Attention has also turned to water. On 14 August, Australia’s water ministers discussed proposed national standards intended to minimise data centre water use, increase the use of recycled and non-drinking water, and limit impacts on households, farmers and other industries.

These developments matter. But as governments work through the regulatory, energy and water challenges associated with data centre growth, it is important to remember that technical standards and approvals are only part of the equation.

Data centres also need the acceptance of the communities around them. As investment in AI data centres accelerates, questions around their social and economic legitimacy are becoming an increasingly important part of the public debate. For AI proponents, gaining and maintaining community trust needs to rank higher on the agenda.

As AI becomes increasingly embedded in Australian workplaces and everyday life, public attention is also turning to the broader impacts of the infrastructure required to support it, particularly the energy and water-intensive data centres underpinning AI growth.

Right now, growth in AI data centres seems unstoppable. As at 31 March 2026, Australia had 162 operational data centres, with a further 90 projects in the development pipeline. In much the same way that railways enabled industrialisation and telecommunications empowered the information age, AI data centres are heralding a new era of human advancement.

But revolutionary advancements come with real impacts.

As those impacts become clearer and communities begin to push back, one thing is certain: AI data centre proponents need to do more than reiterate the game-changing benefits of AI and the inevitable march of progress. Technology providers, developers, operators and government stakeholders must lean into the conversation, answer the hard questions, and address how those impacts will be managed.

This is how social licence is earned.

Social licence – the key to community acceptance

Social licence is critical for major infrastructure projects in Australia. Achieving social licence means the project has a level of acceptance or approval from affected communities and stakeholders.

The Thomson–Boutilier model of social licence provides a useful way to understand how project acceptance develops. It describes social licence as a hierarchy: projects may be rejected or merely tolerated by stakeholders, or they can gain acceptance, then approval and, at the highest level, achieve stakeholder advocacy.

To advance through these levels, projects need to prove their socio-political and economic legitimacy.

AI data centre proponents can build interactional and institutional trust by engaging openly with local communities from the outset. This means clearly explaining what is proposed and why it is needed, being transparent about potential short- and long-term impacts, identifying which decisions remain open to influence, and showing how community concerns will be considered and addressed.

Interactional trust – how AI proponents can build meaningful engagement

Proponents cannot control what people think about AI or how they feel about a new data centre in their neighbourhood. What they can control is how they engage with the people affected by their decisions.

This practical opportunity is identified in the Thomson–Boutilier model as interactional trust.

It means putting aside corporate language, engaging early, listening seriously, explaining constraints and returning with evidence of what has been investigated. It also means continuing to show up even when relationships with the community are strained.

Genuine engagement is not the sole responsibility of a community engagement or corporate affairs team. Project leaders and technical specialists need to participate so questions can be answered, options tested and workable solutions developed.

Communities can tell the difference between a process designed to listen and one designed simply to record that consultation occurred. Not every concern will lead to a changed project. But trust can still grow when a proponent investigates a community request and explains openly why it can or cannot be delivered.

For data centres, this could mean: working with communities to reduce construction traffic, noise and visual impacts; being transparent about water and electricity demands; explaining how climate extremes will be managed; committing to local procurement and construction employment; improving local digital connectivity; or investing in amenities and projects that deliver an enduring community benefit.

Institutional trust – the role of industry and government in AI accountability

The second side of trust is institutional. Do communities trust governments to set appropriate safeguards and hold proponents accountable? Do they trust the data centre and AI industries to act responsibly over the long term?

Individual proponents cannot carry this national conversation alone.

They need to explain the role data centres will play in Australia’s economic and strategic future while honestly acknowledging the impacts.

The Australian Government’s policy response is also becoming more concrete. On 15 July 2026, the Office of AI was established to coordinate the design and legislation of a new Australian AI standard, including mandatory requirements for large AI data centres covering energy and water. This sits alongside the AEMC’s August recommendations on how data centres should source power, manage demand and bear network costs, and the 14 August discussion by water ministers on proposed national standards to minimise water use and increase the use of recycled and non-drinking water. Together, these developments are beginning to translate broad policy principles into more specific expectations for the sector.

Government must establish the rules, protections and public-interest case. Industry must lift transparency and performance across the sector. The proponent must then translate national confidence into local trust in the community where it seeks to operate.

The way forward for AI infrastructure in Australia

AI data centres may be digital infrastructure, but their success will depend on something deeply human: whether the communities asked to host them believe in the benefits and can accept the trade-offs.

To achieve and maintain social licence, proponents need to engage authentically and openly with their communities: explain the need, acknowledge the trade-offs, be clear about what can be influenced, involve the people who can make decisions, and keep showing up, even when discussions are difficult.

By being a force for the community, not against it, AI data centres can build trust and legitimacy and build a path towards achieving and maintaining social licence.