Report

Greening intelligence: Charting the future of sustainable AI

Generative AI is reshaping industries, promising gains in productivity and competitiveness. Yet its rapid ascent has brought a host of challenges, from regulatory uncertainty to mounting energy demands. As firms grapple with the environmental costs of innovation, this briefing paper examines how organisations worldwide are approaching AI sustainability. Drawing on insights from users and suppliers alike, it explores the diverging regional strategies, shifting executive priorities, and the growing pressure for industry-wide standards.

Download report
Executive summary Introduction Opportunities

Executive Summary

Artificial Intelligence (AI) and Generative Artificial Intelligence (GenAI)—algorithmic technology based on large language models (LLM) capable of generating text, images or other data in response to user prompts—can boost productivity and create competitive organisational advantages. As a result, companies in a wide range of industries across the world have embraced AI to seize new opportunities, such as better customer service delivery and improved operational efficiency. Given the rise of GenAI in particular, awareness has risen to executive levels at industry-leading organisations seeking to reap the full benefits of implementation.

However, companies also face challenges in adopting this relatively new technology, including a lack of regulations at the international, national, industry, and organisational levels, which leads to concerns regarding protection of privacy and proprietary information, cybersecurity, and environmental sustainability linked to increased energy usage. Such concerns can hinder the full potential of AI implementation across all industries.

This report looks at current and future AI sustainability from both a supplier and user perspective globally. Research shows that some organisations have embarked on AI sustainability ventures and that a vast majority of executives expect more to be done in the near future. Cost is not a concern, but the roadmap towards AI sustainability is murkier.

The key findings of the research are as follows:

AI sustainability is becoming increasingly important to organisations globally.

Companies recognise the need to balance AI innovation with environmental concerns as the energy demand for AI surges, with a significant increase expected in the perceived importance of AI energy efficiency over the next 12 months.

There is no silver bullet to improve AI sustainability—organisations take diverse approaches based on specific needs.

Reflecting the complex and multifaceted nature of the challenge, companies are implementing various strategies including energy audits, partnerships with efficiency organisations and optimising hardware performance.

Cost is not a primary barrier to AI sustainability efforts.

Organisations are prioritising life-cycle assessments, infrastructure evaluations and real-time energy monitoring over cost considerations when evaluating the sustainability of AI solutions, indicating a willingness to invest in long-term sustainability.

Grid resilience challenges.

Power systems must contend with increasing climate volatility, ageing infrastructure and rising energy demands. Stable electricity supply and power-grid resilience are the top challenges among European survey takers, while energy consumption and carbon footprint is seen as the main barrier among executives from Asia-Pacific.

There are significant regional differences in approaches to AI sustainability.

Executives in North America, Europe and Asia-Pacific have varying views on the role of government regulations in encouraging AI sustainability, highlighting the need for global collaboration and standardised approaches.

The gap between AI users and suppliers in sustainability expectations is widening.

AI users increasingly expect suppliers to have sustainability targets, with a dramatic rise anticipated over the next 12 months, pushing the industry towards more sustainable practices.

AI-driven innovation offers great potential to advance sustainability, but industry obstacles continue to slow its progress.

Edge computing and hardware infrastructure are central to sustainability efforts, yet AI suppliers identify both technological and process innovation as significant barriers to faster uptake, underscoring that industries have yet to fully realise AI’s sustainability benefits.

Introduction

I guess it’s an over simplistic conclusion, but AI is basically making things possible that were not possible before.
Ricardo Vinuesa, associate professor, KTH Royal Institute of Technology, Sweden

The widespread media coverage of ChatGPT in 2023 served as a tipping point for GenAI, raising awareness of the opportunities—and challenges—of AI technology generally. Bloomberg Intelligence, a consultancy, predicts that the global GenAI market alone will grow by over 40% annually, from US$40bn in 2022 to US$1.3trn by 2032.

In the survey conducted for this report, the importance of AI solutions is expected to increase significantly:

47%

of survey takers consider AI-related operations or solutions to be “important” today, rising sharply to,

91%

within 12 months—highlighting the near-universal role AI is set to play across industries.

“Most of the benefits from AI come from new technologies or new resources, new methods that were not possible before,” says Ricardo Vinuesa, associate professor at the KTH Royal Institute of Technology in Sweden. “I guess it’s an over simplistic conclusion, but AI is basically making things possible that were not possible before.”

We’re in a period of chaotic innovation.
Josh Parker, head of sustainability and senior director of corporate sustainability, Nvidia

As rapid technological advancements occur and the ability to process and analyse vast datasets increases, AI is expected to have a significant impact on a vast range of sectors, including logistics, finance, pharmaceuticals, education, technology and government. This impact extends not only to organisations that use AI but also those that supply AI solutions. “We’ve been working on data analytics and AI for more than ten years,” says Prerit Mishra, head of data and analytics at DHL Asia-Pacific. “We also adopted generative AI fairly quickly.”

44% of the companies say that AI energy efficiency is important today but will increase to 78% in the next 12 months.

No silver bullet

The rapid growth of AI is driving up energy consumption and carbon emissions, with IT emissions in consumer products alone expected to triple. Therefore, the importance of AI energy efficiency is expected to increase significantly: only 44% of respondents consider energy efficiency important in their AI-related operations or solutions today, but that figure rises to 78% when looking 12 months ahead from now. Companies are recognising the need to balance AI innovation with sustainability concerns, as the energy demand for AI is surging. However, organisations do not see a “silver bullet” solution. Instead, companies are exploring a mix of strategies to address the complex challenges of balancing AI innovation with environmental concerns.

Organisations have adopted a range of strategies tailored to their operations.

Organisations have adopted a range of strategies tailored to their operations. Alibaba, a China-based global technology company and an e-commerce giant, emphasises its “Five Greens” strategy, which includes green energy, green products, green architecture, green operations and green services.

Alibaba, a China-based global technology company and an e-commerce giant, emphasises its “Five Greens” strategy, which includes green energy, green products, green architecture, green operations and green services.

“We have adopted a holistic approach to sustainability,” says Roger (ZhaoYang) Wang, general manager, Global Data Center, at Alibaba Cloud. “This includes [using] metrics like carbon emissions per computing power and power consumption per computing power to guide our development.” “We’re in a period of chaotic innovation,” says Josh Parker, senior director of corporate sustainability at Nvidia, an American technology company that manufactures important hardware (chips) for AI. “The value per watt is improving rapidly, but we’re still figuring out how to balance efficiency with growing demand.”

Opportunities

There is no one silver bullet — it’s a system-level problem.
Mario Zeng, vice-president, Hardware System Engineering, Microsoft

Users and suppliers are aligned on the importance of AI sustainability: Almost seven in ten respondents (69%) say that the importance of AI sustainability will increase within the next 12 months. For example, AI can assist in developing sustainable offerings that save money, streamline innovation and build new business models.

Meanwhile, consumers and customers continue to rate sustainability as an important purchase criterion, but often lack a clear understanding of what makes a product or service sustainable. In response, companies are using AI to improve communication about sustainable products and propositions to consumers. “When you’re talking about a global network, when you’re talking about the number of deliveries that happen on a daily basis [at a global logistics company like DHL], I think that all of it adds up”, says Mr Mishra (see case study).

“There is no one silver bullet — it’s a systemlevel problem,” says Mario Zeng, vice-president, Hardware System Engineering at Microsoft. “We’re not just scaling within Microsoft — we’re collaborating to shape industry standards.”

To continue reading the report click on the download button below:

Download full report

Video

Charting the future of sustainable AI

Infographic

Greening intelligence: Charting the future of sustainable AI

Discover how AI is transforming our world—and what that means for the planet. This infographic dives into the views of 608 AI users and suppliers, including 362 top executives, uncovering surprising insights, shared goals, and key differences in the journey toward a greener, more sustainable future for AI.

Explore infographic

Learn more about Delta Electronics' approach to AI and sustainability and how our solutions contribute to a smarter, more energy-efficient future.

Visit website