Artificial intelligence (AI) has made a breakthrough in nearly every aspect of our lives – from automating simple tasks like grocery lists to completely transforming industries such as healthcare, manufacturing, transportation, and more. Around the world, organizations are recognizing AI’s potential to drive efficiency and innovation. However, one critical challenge is coming to the forefront: AI’s environmental impact.
From training models to running data centers to the e-waste created, the total footprint of AI is much bigger than many may realize. Research shows that to train large, widely used AI models, the process can emit 600,000+ pounds of CO2e – which is equivalent to 5x the lifetime emissions of an average American car. For a simple 20-50 question conversation with a widely used AI chatbot, it is estimated that the model ‘consumes’ the equivalent of a 500ml water bottle. As AI adoption continues to accelerate, companies must now find ways to balance embracing these technological advancements with meeting and maintaining key sustainability goals.
The Environmental Impact of AI
AI and machine learning technologies are using more energy than ever before. Training models demand an incredible amount of power and according to research, this amount has doubled every 3-4 months since 2012 due to the sheer size, scale, and complexity of the new models being built. It’s important to remember that these models are also not just trained once – they require regular updates to ensure the most relevant information is being used. With this rapid growth, it is estimated that by 2040, emissions from Information and Communication Technologies (ICT) will account for more than 14% of the globe’s total emissions.
New data centers are also being built around the world to power AI use, with a record number of new constructions happening in the U.S. for 2025. According to a report from the International Energy Agency (IEA), the total electricity consumption of these data centers could reach more than 1,000 Terawatt hours by 2026, which is equivalent to the entire electricity consumption of the country of Japan.
The Benefits of AI for Sustainability
Despite its large environmental impact, AI does have practical and powerful use cases for sustainability efforts. Like in many other EHS industries (e.g., healthcare), AI has proven to be incredibly helpful at finding efficiencies and areas of optimization.
When it comes to addressing environmental and sustainable impact, AI can support in understanding key patterns in historical data, conducting scenario analysis, consolidating metrics for more accurate carbon footprint calculation, finding areas to reduce emissions, and much more. The United Nations Environment Programme leverages AI to monitor air quality, specifically tracking methane emissions and aggregating data across geographies.
AI is already revolutionizing sustainability reporting by streamlining data collection, improving accuracy, and enhancing decision-making. Learn more about how AI is transforming sustainability reporting.
AI also has use cases across various industries. From a retail perspective, for example, AI technologies can help companies to optimize their supply chain such as finding ways to tackle transportation challenges, reduce fuel consumption, and leverage more effective delivery options. Retail companies can leverage AI to better manage inventories by analyzing consumer buying habits and patterns and stocking goods accordingly. According to a study by McKinsey, AI can support inventory reductions of 20-50%, cutting both company costs and avoiding any additional product waste.
Balancing Innovation and Sustainability
With this challenge laid out in front of them, how can companies balance these two seemingly competing priorities of embracing the powerful new technologies, while also managing the significant carbon footprint that comes with it?
Like in many other cases, companies should approach this topic with sustainability in mind. Here are some tips:
1. While building an AI strategy, look to assess the total environmental impact, along with the financial and business impacts, and incorporate it within your decision-making process.
2. Before building a net-new model, assess the AI capabilities that exist internally within the organization and find ways to adapt what’s available to the new use cases. As outlined above, training an AI model is one of the most energy-intensive parts of the process, so leveraging what currently exists could be a way to mitigate this.
3. Once models are up and running, ensure that they are being regularly monitored. If a model has reached a point where no progress or improvements are being made, look to stop training it to avoid any unnecessary energy consumption.
4. Switch to renewable energy powered grids and sources within data centers where possible to reduce the amount of energy being used.
5. E-waste continues to take up more and more of our global landfills. Invest in hardware technology that is future-proof and extend the lifecycle of company electronic products where possible.
It is likely that we are just at the beginning stages of truly understanding the power of AI and the ways in which it will change the way we look at the world. While we continue to lean in and get excited about what’s ahead, it is also equally important to know what we are leaving behind.
Sustainable AI Adoption: Cority’s Approach
At Cority, we believe innovation and environmental responsibility must go hand in hand. As we invest in AI-powered tools to help organizations identify and mitigate risk faster, improve decision-making, and streamline sustainability reporting, we’re also mindful of AI’s environmental footprint. Rather than building energy-intensive models from scratch, we focus on embedding smart, lightweight AI applications—like ergonomic risk analysis and real-time safety monitoring—into our platform. We work with partners who share our sustainability values and use responsible development practices, so our customers can harness the power of AI while balancing innovation with long-term climate goals.
With AI becoming increasingly important in sustainability analytics, companies may have high expectations but are unsure about how to balance innovation with the environmental impact of AI itself. Our award-winning team of 70+ sustainability professionals can help you assess the trade-offs between AI adoption and environmental sustainability, and advise on best practices to mitigate AI’s carbon footprint. To connect with an expert, contact us today.
