The AI Sustainability Paradox – And How to Solve It

Sustainability now challenges executives on several immediate fronts—as concerns from investors, consumers, and employees, and as a regulatory issue. Sustainable practices have also become integral to both efficiency gains and long-term business value. But achieving sustainability requires more than a shift in mindset. It also takes a lot of work. Perhaps that’s why, as part of the United Nations Climate Change Conference COP27, the World Economic Forum made the connection between digital technology and corporate sustainability initiatives.[1]

Quite simply, technology can serve as a force multiplier in the effort to create a greener economy and society. Notable gains are already being made across industries and companies using artificial intelligence (AI) to overcome very real challenges and concretely meet sustainability goals.

A recent McKinsey Global Survey found 83 percent of C-suite leaders and investment professionals expect an increase in shareholder value from environmental, social, and governance programs.[2] Add to that, significant shifts in consumer attitudes and behaviors around environmental and ethical values.

Yet, as companies aim to make good on their net-zero pledges and circular economy promises, they must also show sustainability’s impact on their company’s bottom line. This means reporting on how initiatives will improve productivity, reduce costs, and increase revenues. All while measuring progress against benchmarks and over time, transparently, to stay credible and accountable to customers and shareholders. It’s a tall order and a global imperative at this point. 

How AI Can Benefit Sustainability

In AI’s various incarnations—from machine learning to computer vision—the technology is already helping leaders manage their company’s environmental impacts and mitigate climate risks. AI sustainability applications span industries and business needs from energy and agriculture to supply chains, environmental monitoring, and disaster prediction and response. Some recent notable examples include:

Adding Precision to Agriculture: In the agriculture sector, Nature Fresh Farms is one of the largest independent greenhouse produce growers in Canada. Data collected from sensors and high-resolution video at the edge enable growers to create optimal conditions to enhance produce quality and yield. To reduce overall water usage, spoon-like devices under each plants measure how much water doesn’t get used so they can adjust irrigation as needed.

Making Buildings Smarter: Companies in many different industries are using AI to control new smart buildings that make better use of energy. For example, Siemens is helping customers reduce their buildings’ carbon footprints by leveraging edge and AI technologies to address building performance issues in real time. And AI is helping data center operators make predictions on energy demand and supply. It then identifies the best next steps to minimize energy use at each data center while maintaining safety standards.

Tracking Food Waste: Research shows nearly 15 percent of purchased food ends up being wasted. By applying an AI-based computer vision tool, chefs in commercial kitchens can pinpoint waste, cutting costs and contributing to sustainability goals.[3]

Making Renewables Viable: Utilities can take an AI approach to resolving a key issue with solar and wind energy—intermittency. Through data-driven predictive modeling of weather patterns as well as power supply and demand, AI gives utility operators insight to plan for spikes and make adjustments.[4]

Reducing Environmental Impact of Production: Emerson is relying AI running on data collected from sensors on the factory floor to help provide a single pane view of operations. This allows customers to automate operations for increased efficiency and decreased energy use.

Potential Drawbacks and Proactive Solutions
While AI moves companies closer to their sustainability goals, it could also set them back by increasing their energy and resource consumption. A recent study found AI to be an enabler for 79 percent of targets related to the UN’s Sustainable Development Goals, but reported that in 35 percent of targets across all SDGs there may be a negative impact.[5] 

Given this additional constraint, forward-thinking executives are investing in cleaner technologies. Often this means developing models that can run at a responsible scale. It could also include sourcing data centers with renewable energy or running AI workloads during off-peak hours. Of equal importance, working with companies that are making good on their pledges to lower emissions and use sustainable materials. 

Dell Technologies, for instance, has committed to engineering products with energy efficiency and infrastructure optimization as a priority. The company has used close to 400 million pounds of sustainable products and packaging, and reused or recycled more than 2.5 billion pounds of used electronics since 2007. The same concern for sustainability goes into its products to help customers reach their own initiatives. Dell’s PowerEdge servers are 83 percent less energy intensive than servers from 2013[6], and new PowerMax 2500 storage systems provide 80% power savings per terabyte compared with previous models.[7]

Attention is also paid to reducing the heat generated from Dell’s powerful machines to avoid energy wasted cooling the data center. The new layout of the PowerEdge creates air flow channels that release heat quickly, and Dell Direct Liquid Cooling reduces energy costs by up to 45% relative to cooled air and helps extend the life of existing air infrastructure.[8]

Thanks to advances in AI and the hardware needed to run it efficiently, advancing sustainability using AI is no longer a paradox. Companies can now take advantage of digital technology to answer demands from stakeholders while reducing their use of the Earth’s precious resources and maintaining long-term profitability. 

Learn how Dell Technologies is helping our customers drive positive solutions to achieve their sustainability goals through the power of innovative products and services designed to reduce waste, energy use and emissions by visiting To learn more about AI solutions, visit






[6] Based on internal analysis, June 2022

[7] Dell Technologies Claim No. CLM-004322

[8] Based on Dell internal analysis, March 2021, comparing hypothetical air-cooled data center with a cooling PUE of 0.62 to a hybrid data center with a cooling PUE of 0.34. A PUE of 0.21 was assigned to all overhead not attributed to cooling. Operating costs and other factors will cause results to vary. RS Means industry standards cost basis was used to measure typical cooling infrastructure costs and determine projected savings.

IT Leadership