AI and the Environment
Deep Dive: AI and the Environment
Business in the Community launched the Responsible AI Lab: a ground-breaking initiative that brought together leaders from business, government, and academia to co-create a comprehensive blueprint for Responsible AI. From this lab, we’ve established a set of actions that all businesses should focus on, our foundational guidance, and four topical deep dives to more thoroughly explore key issue areas. This deep dive explores AI as it relates to the environment.
Table of Contents
Introduction
AI usage is increasing and becoming more deeply embedded in business operations, which leads to an environmental impact that is hard to ignore. Behind digital services sit energy-intensive data centres, complex supply chains, and growing demand for water, minerals, and infrastructure. At the same time, AI offers powerful tools to improve efficiency, strengthen sustainability data, and accelerate climate solutions. How organisations design, procure, and govern AI systems will determine whether the technology supports environmental goals or undermines them.
Risks and opportunities
The rapid expansion of digital infrastructure is a key driver of AI’s environmental impact. As AI workloads grow1 the demand for electricity and water is increasing2, often concentrated in specific locations where data centres compete with housing, agriculture, and community needs3. At the same time, many organisations rely on outsourced cloud services and have limited visibility over where data is processed, the energy mix used, or the pressure placed on local resources. Without stronger due diligence and clearer procurement standards, these impacts risk being displaced into Scope 3 emissions4 and local ecosystems.
Alongside infrastructure pressures, data quality and fragmented systems continue to limit effective environmental management. Sustainability information is often spread across multiple teams and platforms, making it harder to track progress, meet regulatory expectations, or respond quickly to emerging risks5. While AI can automate reporting and surface patterns across complex datasets, weak inputs or unchecked outputs can lead to misleading conclusions. Over time, over-reliance on automated insights without sufficient human oversight risks weakening decision-making rather than improving it.
These challenges also extend across supply chains. AI depends on specialised hardware6 that relies on critical minerals and labour-intensive processes, frequently linked to environmental degradation7 and human rights risks8. Procurement strategies that prioritise cost or performance alone can reinforce extractive practices and shorten device lifecycles, increasing waste and emissions.
Alongside these risks, there are significant opportunities. AI can strengthen environmental performance when applied deliberately to high-impact areas such as energy management, supply chain optimisation, and climate reporting. It can support faster identification of emissions hotspots, improve water9 and energy efficiency10, and enable better tracking of Scope 3 impacts. Used responsibly, AI can also accelerate innovation in clean energy, materials, and nature protection11, while supporting behaviour change through smarter feedback and nudges12. It can also enable wider circular approaches by extending asset life, improving resource efficiency and reducing waste across manufacturing, logistics, and recycling systems.
Realising these benefits requires a systems approach. Environmental considerations need to be embedded into AI procurement, governance, and design decisions, not treated as an afterthought. Organisations that align AI strategy with sustainability goals, invest in shared data foundations, and collaborate across sectors are better placed to reduce AI’s footprint while using the technology to drive meaningful environmental progress.
Why does this matter for your business?
If your business does not address the environmental impact of AI, digital innovation could undermine your climate and sustainability goals. Your business could inadvertently increase emissions, water use and environmental harm through energy-intensive infrastructure, opaque supply chains and poorly governed data systems. Further, over-reliance on automated sustainability insights also risks misleading your decision-making. Without a systems approach, your organisation may shift environmental impacts out of sight rather than reduce them, damaging your credibility and increasing long-term risk as expectations and regulations evolve.
Actions by maturity level
Adopting
For organisations beginning to consider the environmental implications of AI, where the priority is gaining basic visibility and avoiding unintended environmental and human rights impacts.
Embedding
For organisations seeking to move from awareness to more consistent environmental management as AI becomes embedded in operations and supply chains.
Leading
For organisations taking a proactive approach to reducing AI’s environmental footprint and using AI to strengthen environmental performance.
Transforming
For organisations aiming to shape wider systems, standards, and collaboration to address environmental impacts that cannot be managed by individual organisations alone.
Case studies
Adopting
We have focused on finding case studies/examples from embedding onwards to inspire businesses and showcase what can be done in more advanced stages.
Embedding
Accenture partnered with a North American technology client to implement a cloud-based ESG reporting solution using AI to automate Scope 1, 2, and 3 emissions data collection. The system improved data accuracy, enabled monthly reporting, and reduced manual effort across sustainability teams. By auditing AI-generated outputs and aligning teams around shared data, the organisation embedded AI into environmental reporting and decision-making13.
Dell Technologies embeds environmental and human rights considerations into its technology supply chains through robust mineral sourcing standards. The company requires all in-scope suppliers to complete conflict minerals reporting, conducts third-party audits of smelters and refiners, and integrates results into supplier scorecards. This approach extends accountability beyond direct operations and reflects leadership in responsible procurement for AI-related hardware14.
Leading
Do you think your business is leading in the area of AI and the Environment? If so, we would love to hear from you and share your story here. Contact your Relationship Manager if you are a BITC Member or info@bitc.org.uk.
Transforming
The World Bank’s GovTech Innovation Lab uses AI to strengthen environmental governance and risk assessment across development projects. By applying AI to large-scale datasets, the Lab supports better identification of environmental and climate risks, improves decision-making in public investment, and promotes transparency across countries and sectors. This approach operates at system level, shaping how environmental risks are assessed and managed in public projects, and aligns closely with efforts to co-create accountability frameworks and shared metrics for sustainable development15.
Veolia used AI to optimise water treatment at a data centre in Illinois, cutting water use by 50 per cent through real-time monitoring and automated adjustment. The system doubled water reuse cycles, saving millions of gallons annually and delivering clear environmental benefits in a resource-intensive context. While this example does not reflect policy advocacy or standard setting, it demonstrates how AI can drive transformative environmental outcomes when applied to high-impact operations and aligned with sustainability goals16.
Endnotes
1. IEA, 2024. Electricity 2024: Data Centre Energy Demand Set to Double by 2030. International Energy Agency.
2. IEA, 2022. CO₂ emissions in 2022.
3. Liu, F. H. M., Lai, K. P. Y., Seah, B. & Chow, W. T. L., 2025. Decarbonising digital infrastructure and urban sustainability in the case of data centres. Nature Urban Sustainability, 5, Article 15.
4. Scope 3 emissions refers to all indirect greenhouse emissions that occur in a company’s value chain, both the upstream and downstream of their operations/activities (Carbon Trust 2025).
5. PwC, 2023. How ESG and AI Are Converging. PricewaterhouseCoopers.
6. Environmentally conscious approach. Resources Policy, 78, p.102851.
7. UN DESA, 2025. United Nations Department of Economic and Social Affairs (2025) Harnessing the Potential of Critical Minerals for Sustainable Development.
8. BSR, 2022. Human Rights in the ICT Sector: Supply Chains and Minerals.
9. Veolia, 2024. The water cost of Artificial Intelligence technology. Smart Water Magazine.
10. IEA, 2025, Energy and AI, IEA, Paris https://www.iea.org/reports/energy-and-ai
11. WWF, 2023. AI and Biodiversity Monitoring. World Wildlife Fund.
12. IPCC, 2023. AR6: Sixth Assessment Report (IPCC, 2023).
13. Accenture, 2024. ESG Reporting: From Compliance to Competitive Advantage.
14. Dell Technologies, 2025. Conflict Minerals Report. Dell Technologies Inc.
15. Okahashi, A. & Blanco, C. How is the World Bank using AI and Machine Learning for Better Governance?
16. Veolia, 2024. Artificial Intelligence is Using a Ton of Water. Here’s How to Be More Resourceful.
Explore our foundational guidance and other responsible AI deep dives
AI and Ethics, Governance & Strategy
Building trust through transparent and ethical governance of artificial intelligence.
AI and Employment & Skills
Equipping people with the skills and confidence to thrive in an AI-enabled world.
AI and Diversity & Inclusion
Preventing bias, widening access and ensuring AI supports inclusive workplaces.
AI and Health & Wellbeing
Protecting autonomy, setting healthy digital boundaries and supporting mental wellbeing.
Thank you to our sponsors and contributors
We would like to thank Deloitte and Verizon for sponsoring the Responsible AI framework. We are also grateful to all the organisations, members and academic partners for their generous contributions, insights and expertise, which have meaningfully shaped the development of this framework, including BITC members, Verizon Business, Deloitte, Grant Thornton, Pinsent Masons, and Shoosmiths, Dr Luca Arnaboldi, Dr. Mehreen Ashraf, Emre Kazim, Dr Felicia Liu, Zhuang Ma, Roberta Pierfederici, Dr Daniel Wheatley, Allwyn UK, Cancer Research UK, Good Things Foundation, Macmillan Cancer Support and UKAI.