Deloitte Asks: How Can We Achieve ‘Green AI’?

Share
Deloitte Global's Powering AI study highlights the need for Green AI
A recent report from Deloitte Global explored the climate impacts of AI, highlighting a need to reduce the environmental footprint of this technology

In our current digital age, AI is becoming an integral piece of technology in our everyday lives due to its incredible ability to solve complex problems efficiently and automate many tasks.

However, this powerful piece of technology has a much darker side.

Global professional services network Deloitte recently released a report exploring the environmental impact of AI's rapid expansion.

One of the most significant issues that arose was the global electricity use by data centres, which is expected to triple in the next decade. 

The report brought this significant rise in energy consumption down to the growing computational demands that are being asked from AI. 

As a result of this, Deloitte stresses the need to ensure that AI’s development aligns with global climate objectives. 

Youtube Placeholder

The environmental footprint of AI

Around 0.3% of global greenhouse gas emissions come from data centres - a key foundation for the growth of AI.

In addition to this, energy-intense processors like GPUs and TPUs - which use more energy than traditional computing - are constantly required to train and run AI models, therefore contributing further to the environmental footprint of AI. 

The rapid growth of the integration of AI into industries is particularly dangerous in regions that are still reliant on fossil fuels for power, as this growth results in greater computational workloads and electricity consumption. 

Without implementing regulatory frameworks and fostering industry-wide collaboration, we risk AI diminishing our progress toward sustainable goals. 

Deloitte highlighted the urgent need to adopt Green AI (Credit: deloitte.com)

Deloitte Global’s ‘Powering AI’ report 

Deloitte Global highlights the urgent need for stakeholders to collaborate and craft sustainable solutions to make sure AI development orients with worldwide sustainability goals. 

The study centered on three key questions on the environmental impacts of AI and energy consumption -

  • What are the current trends, key drivers, and regulatory frameworks that will help shape the evolution of AI and the energy consumption of data centres?
  • How might the energy consumption of data centres evolve, and what would the resulting carbon footprint be?
  • What strategies can businesses and governments implement to help mitigate the environmental impact of AI-related energy consumption on climate?

It was shown that data centres consume 380TWh of electricity globally (which is around 1.4% of global electricity use). By 2030, this could hit 1,000TWh and 2,000 TWh in 2050 (around 3% of global energy consumption). 

If AI demand continues to accelerate at its current pace without any efficiency improvements, Deloitte explained how energy demand could exceed 3,500 TWh by 2050.

To mitigate the damaging environmental effects of the growth of AI, the report highlighted how key strategies for sustainable AI must be implemented.

Geoff Tuff, Principal of Deloitte Consulting LLP, Sustainability and Climate Leader for Energy, Resources & Industrials (ERI) and US Hydrogen Practice Leader

Geoff Tuff, Principal of Deloitte Consulting LLP, Sustainability and Climate Leader for Energy, Resources & Industrials (ERI) and US Hydrogen Practice Leader, explained, "The potential of AI to reduce waste and optimise supply chains is great but so is its energy consumption. [..]  

"I haven't yet seen, at scale, the rational conversation about balancing AI model sophistication based on dominant use with energy intensity and trade-offs."

Deloitte suggested a framework which consists of four pillars
  • Efficiency infrastructure design
  • Clean energy adoption
  • Ecosystem approach
  • Research and innovation
Youtube Placeholder

What is green AI?

Deloitte Globa’s ‘Powering AI’ report highlights the need to adopt ‘green AI’.

This concept strives to stabilise rapid technological advancements with environmental responsibility.

By crafting AI technologies that optimise resources, are aligned with sustainability goals and are energy-efficient, we can ensure that the future development of AI contributes positively to our sustainability goals and climate.

Green AI will enhance hardware and data centre operations to minimise energy efficiency, improve transparency by forming standardised reporting on energy use and focus on green energy sources to power AI operations.

Examples of Green AI
  • Green data centres
  • Heat recovery
  • Efficiency cooling systems
  • Sustainable computing
  • Resource optimisation
  • Energy efficiency

Altogether, green AI will support a sustainable digital ecosystem and ensure that its incredible benefits fail to heighten environmental challenges. 


Explore the latest edition of Sustainability Magazine and be part of the conversation at our global conference series, Sustainability LIVE

Discover all our upcoming events and secure your tickets today.


Sustainability Magazine is a BizClik brand

Share

Featured Articles

DNV: Sustainability Tops the Agenda for Food & Drink Sector

According to research by DNV, leaders from the food and drink sector are treating sustainability as top priority, driven by pressures from all sides

New York City's Fifth Avenue Set for a Sustainable Revamp

New York City has plans to make Fifth Avenue a greener, safer and more accessible space with potential to boost the city's sustainability and economy

Shell v Greenpeace: How the Climate Clash Has Ended

In 2023, Shell sued Greenpeace after activists occupied an oil platform off the Canary Islands, leading to a multimillion-dollar legal dispute

HubSpot: Democratising Sustainability and Depoliticising ESG

Sustainability

Orange & La Poste: Recycling Olympic Network Technology

Tech & AI

Knight Frank: Are All Carbon Credits Created Equal?

Net Zero