Part 2 of our 5-part series.

Artificial intelligence (AI) is rapidly transforming businesses, promising enhanced efficiency, innovation, and growth. However, beneath AI’s transformative potential lies significant environmental and social concerns, particularly around water usage and rare earth mineral extraction. Addressing these ESG considerations is not merely a compliance exercise but an essential aspect of responsible leadership and future-oriented corporate strategy. 

Assessing AI’s water dependency in a resource-constrained world 

Data centres are central to AI operations and are among the are among the most intensive consumers of water resources in the ICT sector, particularly when compared to traditional enterprise data servers and other industrial digital infrastructure, due to their high-density compute loads and cooling requirements. Conventional methods such as evaporative and HVAC cooling systems contribute substantially to global water stress, particularly in regions already grappling with scarcity. For example, major technology hubs hosting large-scale data face growing scrutiny over their water usage impacts prompting stakeholders and investors alike to demand greater accountability.¹

To proactively address these concerns, industry leaders are adopting innovative cooling solutions. Liquid immersion cooling, where servers are fully submerged in non-conductive liquids, offers significant reductions in water use compared to traditional cooling methods. Likewise, free-air cooling systems leveraging external ambient air substantially curtail the need for water-intensive cooling. Incorporating these advanced technologies into strategic ESG planning provides companies with competitive differentiation and positions them as proactive stewards of critical natural resources. 

CDP’s enhanced emphasis on water management for the 2025 reporting year underscores the escalating importance of water stewardship across operations as well as their supply chains. Companies will be required to deliver comprehensive disclosures on water use, management practices, and mitigation strategies, further aligning ESG performance with investor expectations and global sustainability benchmarks. 

Rare earth and critical minerals: Ethical implications and ESG supply chain transparency 

AI’s indirect reliance on rare earth and critical minerals such as lithium, cobalt, and neodymium creates pronounced environmental and social risks. These materials support a wide range of critical hardware from batteries, high-efficiency magnets, to the micro-sized transistors and connectors inside chips and accelerators. The extraction and processing of these critical minerals involve significant ecological damage – polluting water resources, degrading habitats, and generating toxic waste. Socially, the mining of these materials frequently intersects with severe human rights violations, including forced labour, unsafe working conditions, and community displacement. 

The accelerated global adoption of AI is set to dramatically increase the demand for these minerals, elevating associated ESG risks. Transparent, accountable supply chains are no longer optional – they are imperative. Organisations committed to ESG excellence must undertake rigorous sustainability audits, align with international frameworks such as the OECD Guidelines for Responsible Mineral Supply Chains and the UN Global Compact, and integrate blockchain-based traceability systems to assure ethical sourcing practices. Such transparency not only mitigates risks but also fortifies brand reputation and enhances investor confidence. 

Advancing ESG measurement methodologies for AI 

The complexity of quantifying AI’s ESG impacts necessitates robust methodologies. Lifecycle Assessments (LCAs) offer comprehensive tools to evaluate environmental impacts across the entire lifecycle of AI infrastructure – from raw material extraction through operational use to end-of-life disposal or recycling. LCAs, alongside regular ESG-focused supplier audits, enable companies to generate actionable insights and align their practices with globally recognised reporting frameworks such as GRI and SASB. 

Improving measurement methodologies and transparency in reporting allows businesses to more effectively manage ESG risks, adapt to regulatory expectations, and demonstrate leadership in corporate responsibility. 

Innovative strategies for sustainable AI implementation 

Addressing the intertwined challenges of water usage and mineral extraction demands strategic innovation. Sustainable sourcing from certified suppliers adhering to responsible mining standards like The Initiative for Responsible Mining Assurance and Responsible Mining Initiative are foundational. Complementing this approach with circular economy practices, such as re-using and recycling metals and minerals from technologies which have reached the end of their usable life, significantly reduces reliance on primary extraction. 

Further, investing in energy – and resource – efficient AI technologies through model optimisation techniques – such as pruning, transfer learning, and fine-tuning – reduces computational demands, lowering the overall environmental impact. Concurrent research and development into alternative, less environmentally impactful materials also represent strategic opportunities for forward-thinking businesses. 

AI’s immense potential can only be realised sustainably if organisations proactively tackle ESG challenges related to water and rare earth minerals. Companies embracing advanced technical solutions, rigorous ESG reporting, and ethical supply chain practices will distinguish themselves as visionary leaders in the evolving digital economy, creating long-term value for both stakeholders and society at large. 

In our next article in this AI series, we examine the social implications of AI, from algorithmic bias to automation and how businesses can integrate ethical governance into their digital transformation. 

Author: Harry Freeman, Consultant, Simply Sustainable

1. We now know how much water Google’s Oregon data centers use, after city drops lawsuit against journalists – DCD