Artificial intelligence is often discussed through the lens of software, algorithms and productivity gains. Yet behind every AI model sits a rapidly expanding physical infrastructure network: data centres, semiconductors, electricity grids, cooling systems, water supplies and the critical minerals required to manufacture advanced computing equipment. As AI adoption accelerates, this infrastructure is becoming one of the most important investment themes of the decade. For investors, understanding both the opportunities and risks emerging from this transition is essential.
Much like electrification, railways and the internet before it, AI requires enormous enabling infrastructure. Every AI-generated image, document, recommendation or decision relies on physical assets somewhere in the world. The AI value chain extends from quartz, silicon and critical minerals through semiconductor manufacturing and data centres to foundation models and downstream applications. Importantly, risks and opportunities exist across every layer of this value chain rather than solely in the technology companies developing AI models.
From an investment perspective, the opportunity is significant. Demand for high-performance semiconductors, cloud infrastructure, data centre capacity, renewable energy, grid upgrades and digital connectivity is increasing rapidly. AI is also creating productivity improvements across sectors, enabling companies to automate routine processes, improve decision-making and develop new products and services. Companies that successfully harness AI may generate material competitive advantages.
However, focusing solely on the growth opportunity risks overlooking a critical reality: AI infrastructure is increasingly becoming a sustainability and governance issue.
Data centres sit at the centre of this challenge. They are the physical backbone of AI and represent one of the largest sources of AI-related environmental impact. Advanced AI systems require substantial computing power, which translates into significant electricity demand. Cooling systems often require large volumes of water, while backup systems, construction materials and equipment manufacturing create additional environmental footprints. AI infrastructure is therefore deeply connected to energy systems, water availability, climate goals and community expectations.
For investors with exposure to infrastructure, property, utilities and listed equities, these issues are no longer theoretical. Questions around energy sourcing, water consumption, community impacts and social licence increasingly influence the long-term value and resilience of AI-related assets. Communities hosting large-scale data centres are becoming more focused on local impacts such as water use, land use, energy availability, noise and infrastructure requirements. Regulators are also beginning to respond. Governments are introducing new expectations for data centres and AI infrastructure developers, while overseas markets are moving towards mandatory disclosure of energy efficiency, water efficiency and sustainability performance.
This emerging regulatory landscape reflects a broader trend. AI governance is evolving from a voluntary principle-based approach towards more formal and enforceable frameworks. Investors should expect increasing scrutiny of not only how AI systems are designed and deployed, but also the infrastructure that enables them. Companies that fail to anticipate these developments may face rising operational costs, planning delays, regulatory intervention or reputational damage.
The challenge for investors is therefore twofold. First, they must understand their exposure to AI infrastructure throughout their portfolios. Exposure is not limited to data centre operators or technology companies. It can also include utilities supplying electricity, property owners hosting facilities, critical minerals producers, semiconductor manufacturers and companies heavily dependent on third-party computing services. AI-related risks can propagate through the value chain, making visibility into suppliers and infrastructure providers increasingly important.
Second, investors have an important stewardship role to play.
Responsible investment has always been about ensuring that companies manage material risks and opportunities effectively. AI should be no different. Investors should expect companies to identify and manage the environmental, social and governance impacts associated with AI systems and the infrastructure supporting them. This includes visible accountability, robust risk management processes and transparent disclosure.
For AI infrastructure providers, this means demonstrating how energy and water consumption are measured and managed, how facilities are designed to improve efficiency, and how local communities are engaged throughout project development and operation. Disclosure of metrics such as power and water utilisation can help investors assess performance and identify leaders and laggards. Investors should also seek evidence that companies are considering biodiversity, climate impacts, resilience and long-term resource availability in infrastructure planning decisions
Investors should also broaden their perspective beyond environmental impacts alone. AI infrastructure creates social and governance considerations that are equally important. The extraction of critical minerals raises questions around labour standards and supply chain integrity. Concentration within semiconductor manufacturing and AI infrastructure creates resilience and systemic risk concerns. Dependence on a handful of providers may expose businesses, markets and economies to operational disruptions or geopolitical tensions.
At the same time, responsible investors should recognise that AI can support sustainability outcomes. AI has the potential to improve energy efficiency, optimise resource use, accelerate scientific research and support climate transition efforts across many industries. The objective is therefore not to slow innovation, but to help ensure that innovation occurs within appropriate governance frameworks and with sufficient attention to long-term sustainability considerations.
Ultimately, AI infrastructure represents one of the defining investment challenges and opportunities of the coming decade. The winners are unlikely to be determined solely by who can build the largest models or the fastest chips. Increasingly, success will depend on who can build resilient, efficient and socially accepted infrastructure capable of supporting AI growth in a resource-constrained world.
For long-term investors, the question is not whether AI infrastructure will be important. The question is whether the companies benefiting from this growth are managing the accompanying environmental, social and governance risks with the discipline, transparency and accountability necessary to create sustainable value. Responsible investors have a critical role in helping ensure that the answer is yes.
Key investor actions on AI infrastructure and responsible investment
1. Map portfolio exposure across the ai value chain
Investors should move beyond viewing AI as purely a technology issue and assess exposure across the full AI value chain, including:
- Critical minerals and raw materials
- Semiconductor manufacturers
- Data centre operators
- Utilities and energy providers
- Cloud and infrastructure providers
- AI developers and deployers
AI-related risks often sit outside traditional technology holdings and can be embedded throughout listed, infrastructure, property and private market portfolios.
2. Engage data centre operators on sustainability performance
Investors should seek evidence that data centre operators are:
- Managing energy consumption efficiently
- Minimising water use and water stress impacts
- Sourcing renewable energy
- Considering biodiversity and land-use impacts
- Managing community impacts and maintaining social licence
Key metrics to encourage include Power Usage Effectiveness (PUE) and Water Usage Effectiveness (WUE).
3. Assess AI infrastructure resilience
Investors should evaluate:
- Reliance on concentrated suppliers and infrastructure providers
- Exposure to semiconductor supply chain bottlenecks
- Geopolitical and regulatory risks
- Availability of long-term energy and water resources
- Operational resilience of critical AI infrastructure
Concentration and systemic risk are emerging material investment considerations across the AI ecosystem.
4. Expect strong governance and accountability
Investors should engage companies to ensure:
- Clear board or senior management accountability for AI
- AI risks are incorporated into enterprise risk management
- Appropriate oversight and monitoring processes exist
- AI governance is treated with the same rigour as other material business risks
Good governance is increasingly becoming a differentiator between companies creating sustainable value and those exposing investors to future liabilities.
5. Demand better disclosure
Investors should encourage reporting on:
- AI-related energy and water consumption
- AI governance frameworks
- Material AI incidents
- Supply chain management practices
- Workforce impacts and transition planning
- Environmental impacts associated with AI operations
Transparent disclosure supports both investment decision-making and stewardship activities.
6. Consider the entire ESG footprint of AI
Responsible investors should assess:
- Environmental impacts (energy, water, emissions)
- Labour conditions in AI supply chains
- Data governance and privacy risks
- Bias and discrimination risks
- Cybersecurity and system resilience
- Social and community impacts
AI infrastructure should be viewed through the same ESG lens applied to other major infrastructure investments.
7. Prepare for rapid regulatory change
Investors should monitor evolving AI and data centre regulation, including:
- AI governance requirements
- Environmental disclosure obligations
- Critical infrastructure regulations
- Water and energy planning frameworks
- Emerging AI-specific legislation
Companies that anticipate regulatory developments are likely to be better positioned than those reacting after rules are introduced.
8. Support AI's positive contribution
Responsible investors should not simply focus on risk mitigation. They should also encourage companies to deploy AI in ways that:
- Improve resource efficiency
- Accelerate climate solutions
- Enhance productivity
- Support innovation and economic growth
The objective is responsible scaling of AI, not restricting innovation.
A simple stewardship question set
When engaging companies exposed to AI infrastructure, investors could ask:
- How is AI changing your energy and water footprint?
- What AI-specific risks are overseen by the board?
- How do you assess and manage AI supply chain risks?
- What metrics do you disclose on AI environmental impacts?
- How are local communities considered in infrastructure development decisions?
- How resilient is your AI strategy to supplier, regulatory and resource constraints?
These questions help shift conversations from AI hype to the governance, sustainability and resilience factors that ultimately drive long-term investment outcomes.
Head of Responsible Investments
,
Aware Super
Liza has over 24 years’ experience in the superannuation sector and is a specialist and passionate advocate for responsible investments and sustainable finance. As Head of Responsible Investments at Aware Super, she has led the development and implementation of the Fund’s responsible investment policies, the execution of the Climate Change Strategy and also manages the ESG policy implementation including manager and asset class ESG reviews.
Liza joined Aware Super through its merger partner Health Super in December 2006 as an analyst in the Compliance, Legal and Risk Team. Before joining Aware Super, Liza held various roles at Mercer Legal where her primary focus was on trustee education and corporate secretarial duties.
Liza represents Aware Super on a number of working groups and committees including Investors Against Slavery & Trafficking APAC (IAST-APAC); ESG Research Australia; the Australian Sustainable Finance Institute (ASFI); 40:40 Vision; the Australian Council of Superannuation Investors (ACSI) and the UN convened Global Investors for Sustainable Development Alliance (GISD).
Liza holds a Post Graduate Diploma of Applied Finance at Kaplan Education and is a Graduate of AICD.
.avif)
.avif)
