Businesses worldwide face a critical juncture between artificial intelligence (AI) and environmental, social, and governance (ESG) goals. The burgeoning energy demands of AI, particularly large language models (LLMs) like ChatGPT, pose a significant paradox: how can we harness AI’s potential without compromising our commitment to environmental sustainability? This question is not just rhetorical but a strategic dilemma that businesses must navigate to achieve responsible and sustainable growth. Kush Kahadugoda from Oracle Consulting Australia and New Zealand explores the issues with RMIT University’s Say Yen Teoh, Yiliao Song and Yee Ling Boo.
Key insights
- AI’s energy demands pose a challenge to ESG goals, but solutions exist. Data centers and large language models consume significant energy, but renewable energy sources, smaller AI models, and algorithmic efficiency can mitigate the impact.
- AI can be a powerful tool for ESG as a strategic corporate asset. AI enables more robust reporting, modelling and decision making on ESG. Integrate ESG into your core business strategy to attract investors, enhance brand reputation, and unlock new business opportunities.
- ESG is no longer optional. As regulatory regimes tighten and the reputational and legal ramifications of greenwashing intensify, organisations are increasingly required to address these issues – they’re transforming from “nice to have” to “must have”.
- Practical AI applications are already creating positive social and environmental impact. Examples include data analysis for sustainable investing, operational efficiency improvements, and forest management.
- Responsible innovation is key to navigating the AI-ESG intersection. Businesses and investors need to consider their digital readiness, technology choices, and partnerships to ensure sustainability is prioritized.
- Collaboration between sustainability team(s) and AI/ technology team(s) is vital. The key staff members and leaders responsible for progressing AI and ESG agendas for organizations should prioritise time for collaboration and leveraging each other.
Paradox of AI’s energy demands vs ESG goals
AI’s global energy consumption is projected to rival the annual electricity use of countries like Sweden or Argentina by 2027 (Zhang, 2023). That presents a formidable challenge to ESG goals. Yet, this challenge also offers an opportunity to redefine the role of AI in driving sustainable business practices.
The question we face is not just about readiness but also about necessity. While AI holds promise, it is energy-intensive, i.e., 1-1.5% of current global electricity use is from data centers (Leffer, 2023). Further, with large language models (LLMs) such as GPT-4, LaMDA, ERNIE 3.0, LLaMA, Claude etc dominating the AI landscape, some studies have projected their potential electricity consumption to reach up to 134 terawatt-hours (TWh) by 2027 (Zhang, 2023), similar to the energy consumption of Sweden or Argentina.
These energy demands pose a paradox. Can we use AI to further ESG goals without compromising environmental sustainability? This delicate balance is where the future of responsible business pivots.
“The potential of AI to help address climate change is enormous. We have the ability to make better predictions, analyse vast amounts of data, and create models to optimize resource usage.” – Professor Andrew Ng, Stanford University
Practical AI applications for greater good
AI’s applications in various business sectors offer glimpses into its potential for societal good. For example, AI-powered early detection of equipment failures can prevent unnecessary energy consumption and reduce carbon emissions, making businesses more sustainable and cost-effective. This is not just about technology for technology’s sake. It is about harnessing AI for the greater good, creating a win-win for businesses and society (Morgan Stanley, 2023). From enhancing data analysis for sustainable investing to improving operational efficiency, AI is poised to become a cornerstone of responsible business practices.
The art of the possible
AI has the potential to be a powerful tool in the fight against climate change by helping us reduce global greenhouse gas emissions.
Optimizing Energy Sytstems
- Smart grids: AI can manage energy demand and integrate renewable energy sources into the grid efficiently, reducing reliance on fossil fuels.
- Predictive maintenance: AI can predict machine failures and optimize maintenance schedules, improving efficiency and reducing energy consumption in buildings and industries.
- Demand-side management: AI can analyze user data to optimize energy consumption patterns, encouraging responsible energy use.
Decarbonizing transportation
- Route optimization: AI can optimize transportation routes for vehicles, including delivery trucks and public transportation, reducing fuel consumption and emissions.
- Traffic management: AI can analyze traffic patterns and signal timings to reduce congestion and improve fuel efficiency.
- Autonomous vehicles: While still in development, autonomous vehicles have the potential to operate more efficiently and safely than human drivers, potentially reducing emissions.
Improving manufacturing and supply chains
- Material optimization: AI can analyze product design and predict material needs, minimizing waste and associated emissions.
- Supply chain optimization: AI can optimize transport routes and logistics, reducing transportation emissions and energy consumption.
- Predictive maintenance: Similar to energy systems, AI can predict equipment failures in factories, preventing unnecessary production shutdowns and optimising energy use.
Other applications
- Precision agriculture: AI can optimize fertilizer and water usage in agriculture, reducing emissions from agricultural practices.
- Carbon capture and storage: AI can improve the efficiency of carbon capture and storage technologies, removing existing carbon dioxide from the atmosphere.
- Financial risk assessment: AI can help investors identify and prioritize investments in clean energy and sustainable technologies.
Sources: Boston Consulting Group 2023; United Nations 2023; World Economic Forum 2024
Oracle’s role
On sustainability, Oracle is action oriented, and outcome driven. Since 2020, Oracle has achieved a 47 per cent decrease in total emissions in our operations, and have over 80 per cent of renewable energy use in Oracle Cloud. Oracle aims to provide our customers globally with a zero-emissions cloud by 2025, halve our value chain emissions by 2030 and achieve net zero by 2050.
Oracle products and services help customers on their sustainability journey—including through solutions for energy and water, more sustainable supply chains, and better reporting on environmental initiatives.Oracle’s integration of AI with ESG initiatives has significantly enhanced data analytics, risk assessment investment strategies, and compliance in various sectors (Hyde, 2023; Oracle and Savanta, 2022).
While what we do ourselves is important, the work we do for customers has an impact many times greater.
Questions for reflection
With the backdrop of the global energy crisis and the widespread digitisation and automation with advanced AI models, data centres are leveraging renewable energy solutions to mitigate the environmental impact (Swallow, 2023). Similarly, to reduce the AI carbon footprints, researchers are doing more with less by tuning to small AI models or “small language models” in which the computational efficiency and cost-effectiveness are redefined (Saenko, 2020; Abbas, 2023).
For businesses and investors, it is essential to reflect on a few key areas: the level of digital readiness, the commitment to investing in relevant technologies, such as renewable data centres and small AI models, and the strategic importance of integrating AI into ESG initiatives. It is essential to consider technology and data centre operators on their way to achieving zero emissions. An example of this would be Oracle’s commitment to providing its customers globally with a zero-emissions cloud by 2025 – this translates to using 100% renewable energy for all operations, including Oracle Cloud. These reflections are crucial in determining the preparedness and potential for success in leveraging AI for sustainable and responsible business practices.
Where to from here?
As we navigate the intersection of AI and ESG, we must foster a mindset of responsible innovation, balancing technological advancement with sustainability and ethical considerations. While complex, this journey steers us towards a more sustainable future if we steer the journey ahead carefully. Whether you are a journey starter, mid-way through, or a leader in the field- each step taken is crucial in shaping your organisation’s success.
For journey starters: Focus on building a strong digital foundation, essential partnerships, and upskilling the workforce.
- Initiate a digital audit to assess your current infrastructure and identify areas for improvement
- Partner with education institutes and organisations for skills development workshops in AI and sustainability principles
- Explore cloud platforms for access to renewable energy and energy-efficient data centers
For organizations taking next steps: Emphasise scaling data analytics capabilities, seeking advanced training, and exploring win-win partnerships for successful AI-ESG integration.
- Utilise AI-powered data analytics tools to optimise resource allocation and reduce your environmental footprint.
- Invest in training programs to equip your workforce with the skills needed to leverage AI responsibly
- Partner with industry leaders to develop and pilot innovative AI solutions for specific sustainability challenge
For leaders in the field: Encourage continued innovation, setting benchmarks, collaborating with research partners, and showcasing success stories to inspire others.
- Advocate for industry-wide standards and best practices for responsible AI development and deployment
- Share your success stories and lessons learned to inspire and guide others on their AI-ESG journeys
- Collaborate with academic institutions and technology providers to accelerate research and development in sustainable AI solutions
Conclusion: Fostering responsible innovation
As businesses navigate the AI-ESG nexus, fostering a mindset of responsible innovation is paramount. By leveraging AI responsibly, we can collectively address environmental challenges, social dilemmas, and governance paradoxes. Our concluding viewpoint for this includes:
- Don’t be deterred by AI’s energy demands, but address them head-on. Explore renewable energy solutions, invest in energy-efficient models, and collaborate with technology providers committed to sustainability
- AI can be a powerful tool for ESG as a strategic corporate asset. AI enables more robust reporting, modelling and decision making on ESG. Integrate ESG into your core business strategy to attract investors, enhance brand reputation, and unlock new business opportunities
- ESG is no longer optional. As regulatory regimes tighten and the reputational and legal ramifications of greenwashing intensify, organisations are increasingly required to address these issues – they’re transforming from “nice to have” to “must have”
- Start small and scale strategically. Begin with foundational digital transformation and workforce upskilling, then progress to advanced data analytics and cutting-edge AI applications
- Collaboration is key. Partner with academic institutions, industry leaders, and technology providers to accelerate innovation, share best practices, and drive industry-wide change
This journey, while complex, steers us towards a sustainable future if navigated with care. Collaborations with academic and industry leaders, such as RMIT and Oracle, are crucial in exploring and embarking on a journey to create a lasting impact.
For more information on the potential to innovate in your organisation where AI and ESG intersect, contact Oracle Consulting ANZ Practice Director for Continuous Improvement & Innovation Kush Kahadugoda. Oracle Consulting ANZ works with a range of strategic partners in AI and ESG including RMIT University. More information on Oracle AI and Sustainability solutions can also be found on the Oracle Sustainability Portal and Oracle AI website.
Authors
Kush Kahadugoda, Consulting Practice Director for Continuous Improvement & Innovation at Oracle Consulting ANZ
Dr Say Yen Teoh, Senior Lecturer, School of Accounting, Information Systems, and Supply Chain, Department of Information Systems and Business Analytics, Enterprise AI and Data Analytics Hub, RMIT University, Wurundjeri Country
Dr Yiliao Song, Research Fellow, School of Accounting, Information Systems, and Supply Chain, Department of Information Systems and Business Analytics, Enterprise AI and Data Analytics Hub, RMIT University, Wurundjeri Country
Dr Yee Ling Boo, Senior Lecturer, School of Accounting, Information Systems, and Supply Chain, Department of Information Systems and Business Analytics, Enterprise AI and Data Analytics Hub, RMIT University, Wurundjeri Country
References
Zhang, H. (2023) AI’s carbon footprint can be managed, say computer science professors, viewed 18 January 2024: https://yaledailynews.com/blog/2023/11/01/ais-carbon-footprint-can-be-managed-computer-science-professors-say/
Swallow, T. (2023) Top 10: Green Energy Data Centres, viewed 18 January 2024: https://energydigital.com/top10/top-10-green-energy-data-centres
Abbas, A. (2023) Small But Mighty: Small Language Models Breakthroughs in the Era of Dominant Large Language Models, viewed 18 January 2024: https://www.unite.ai/small-but-mighty-small-language-models-breakthroughs-in-the-era-of-dominant-large-language-models/
Saenko, K. (2020) It takes a lot of energy for machines to learn – here’s why AI is so power-hungry, viewed 18 January 2024: https://theconversation.com/it-takes-a-lot-of-energy-for-machines-to-learn-heres-why-ai-is-so-power-hungry-151825
Leffer, L. (2023) The AI Boom Could Use a Shocking Amount of Electricity, viewed 16 January 2024: https://www.scientificamerican.com/article/the-ai-boom-could-use-a-shocking-amount-of-electricity/
Yeung, E. (2023) How ESG and value investing can complement one another, viewed 21 September 2023: https://www.investordaily.com.au/analysis/53055-how-esg-and-value-investing-can-complement-one-another
Muraoka, M. (2023) NTT sustainable smart city effort is developing tools for well-being, viewed 25 Dec 2023 https://sustainable.japantimes.com/esg/175.
Morgan Stanley (2023) How AI can bolster sustainable investing, viewed 19 November 2023 https://www.morganstanley.com/ideas/ai-sustainable-investing-use-potential#:~:text=AI%20applications%20for%20sustainable%20investing,privacy%2C%20reliability%20and%20model%20bias
Hyde, G (2023) The power of AI for ESG, viewed 10 November 2023 https://www.fintechfutures.com/2023/06/the-power-of-ai-for-esg/
Oracle and Savanta (2022) 2022 ESG Global Study, viewed 25 December 2023 https://www.oracle.com/au/a/ocom/docs/applications/esg-study-no-planet-b-report.pdf
How AI Can Speed Climate Action, viewed 12 March 2024, https://www.bcg.com/publications/2023/how-ai-can-speedup-climate-action
Explainer: How AI helps combat climate change, viewed 12 March 2024, https://news.un.org/en/story/2023/11/1143187
9 ways AI is helping tackle climate change, viewed 12 March 2024, https://www.weforum.org/agenda/2024/02/ai-combat-climate-change/