Too often, matters of innovation are positioned as somehow opposed to sustainability, as though a choice has to be made: protect the planet or protect the bottom line. The truth is, technological advancements can help drive efforts to improve sustainable business practices. By optimizing operations and cutting back on resource-intensive practices, technology can not only make an organization more cost-effective but also advance their sustainability journey.

IBM has been hard at work utilizing innovations in artificial intelligence and generative AI to help organizations find solutions to accelerate sustainability efforts — including several that debuted at Climate Week NYC earlier this week.

We sat down with three of IBM’s sustainability experts to discuss the role of AI in helping organizations improve operations, identify where there is room for innovation, and explore how the environmental costs of these new developments can be accounted for.

Q: What role does artificial intelligence play in the future of sustainability?

Oday Abbosh, IBM Consulting Global Sustainability Leader: AI is not just a tool, but a catalyst for change. It can help organizations make informed decisions to reduce their environmental footprint and build a more sustainable future. Specifically, AI has the potential to transform the way companies approach their sustainability strategy. According to a recent IBM Institute for Business Value study, spending on sustainability reporting exceeds spending on sustainability innovation by 43%. AI can help rebalance these dollars toward innovation, by streamlining data collection and enabling organizations to take more targeted actions that allow them to drive real sustainability progress.

Kendra DeKeyrel, Vice President ESG and Asset Management Product Leader at IBM: AI is crucial to the future of sustainable business practices, and executives know this: According to IBM’s latest State of Sustainability Readiness report, nine out of 10 business leaders surveyed agreed that AI will help achieve their sustainability goals. Yet there’s still work to be done: IBM also found 56% of organizations are not actively using AI in this manner.

Christina Shim, IBM Chief Sustainability Officer: For us, everything starts with data. Data is central to knowing how much energy you are consuming or what is being wasted. Data creates a baseline that underpins every goal, and is the source from which you can determine your current impact, track progress and implement adjustments. Once you have the right data, implementing technologies like AI can be one of the most powerful tools for a sustainability transition.

Q: As the effects of climate change continue to intensify, resiliency will become an important part of mitigating those impacts and continuing operations without disruption. What role can AI play in improving resiliency?

DeKeyrel: AI is a key tool for building climate resiliency, and business leaders urgently need tools: Only half of those surveyed in IBM’s recent State of Sustainability Readiness Report feel fully prepared to address a range of climate risks. To adapt their physical assets, infrastructure, and natural resources to extreme weather and related risks, organizations must make sense of vast environmental data sets. It could take humans months to label and extract insights from relevant climate and geospatial data — but AI shortens that time frame significantly. For example, IBM Environmental Intelligence uses AI and APIs to integrate complex, diverse data sets and quickly extract insights for improved resiliency. EI has a wide range of applications — from predicting storm and tide risks due to sea level change, to tracking biomass and deforestation trends for conservation efforts.

Shim: Climate change affects people in many ways. It’s leading to increased flooding, causing heat stroke and destroying farms and livelihoods. Insurance is becoming unaffordable. Construction delays due to extreme weather are already costing billions globally. We care because it’s affecting our daily lives, no matter where we live. But let’s focus on solutions. AI can significantly enhance resiliency in the face of climate change by analyzing data to predict extreme weather events, natural disasters, and other climate-related risks with greater accuracy and lead time.

For example, with NASA, we released a model that goes beyond AI forecasting to offer new insights for scientists, developers and businesses about both short- and long-term weather and climate conditions. This is one of the most advanced AI models for climate and weather, and it is now available on the open-source platform Hugging Face, which means more eyes on the code and more brain power on the problems.Foundation models can also help inform efforts to reduce “heat island effects.” IBM and the Mohamed Bin Zayed University of Artificial Intelligence are pioneering the application of foundation models to map urban heat islands – areas with significantly higher temperatures compared to surrounding locations in Abu Dhabi. To date, the model has informed efforts that have succeeded in a reduction of heat island effects in the region by more than three degrees Celsius.

Q: In what other areas do you believe the technology can have the largest impact compared to current methods and technologies?

DeKeyrel: I am already seeing our clients benefit from this. For example, Ford Motor Company was seeking to reduce defects and downtime and create a more sustainable manufacturing process. Ford deployed IBM Maximo Visual Inspection (MVI), which not only improved sustainability, but also vehicle quality. As a result, Ford expanded the use of MVI and its AI capabilities to additional plants and presented IBM with its prestigious IT Innovation Award. In Switzerland and the Netherlands, DSM-firmenich Animal Nutrition & Health was eager to leverage data and AI to improve their business’s sustainability and their animals’ health. DSM-Firmenich used IBM Environmental Intelligence Suite and its powerful algorithms to predict and prevent grain contamination, saving Europe’s agricultural industry millions of Euros each year.

Shim: One of our clients uses AI to orchestrate maintenance for more than 100,000 assets across a LEED Platinum-certified neighborhood, which has 94 buildings, including Riyadh’s tallest skyscraper. A Brazilian renewable power producer used Environmental Intelligence and AI models to improve wind forecasts by 15% and solar forecasts by 30%, allowing them to plan better and ensure a resilient and reliable electric grid. At IBM, we use Envizi, an AI-powered solution, to track and analyze our energy data within a single tool across 600 locations. It enables IBM to track and reduce greenhouse gas emissions, energy consumption and water usage efficiently, while also optimizing facilities management and workplace experiences.

DeKeyrel: Ikano Group is a multinational conglomerate with businesses across industries and multiple continents. For this reason, organizing the vast amount of data needed to meet different reporting standards is crucial for their sustainability journey. When looking to prepare for the new EU Corporate Sustainability Reporting Directive, Ikano group tapped in IBM Envizi ESG Suite. They can now easily capture and track more than 15,000 different data types for CSRD reporting thanks to the platform, saving thousands of manhours in the process.

Abbosh: We also see great possibilities in AI for the circular economy. AI can help uncover ways to use resources more efficiently and create cost savings on raw materials, which often constitute a large portion of production expenses. For example, generative AI can discover and recommend materials for replacing PFAS (similar to synthetic chemicals) at manufacturing sites. This will help us to explore the ways in which technologies are helping with circularity, advancing material discovery and finding alternatives to plastics and PFAS.

Q: Does generative AI have a role in this process? How does it compare to current data processing techniques?

DeKeyrel: Gen AI can supercharge all these processes by “filling in the gaps” and modeling data that traditional AI cannot. This allows businesses to go one step further and tackle new use cases especially for Asset Lifecycle Management, like generating asset failure code recommendations.

Abbosh: Three in four executives say manually processed data holds back sustainability reporting, hindering both business and sustainability outcomes. AI can help bring those insights to the surface. Most clients, regardless of the size of the company, have sustainability teams that are stretched, trying to manually chase data instead of focusing on what the data is saying. Generative AI can unlock productivity potential. As an example, instead of sustainability teams manually collecting and reviewing paper fuel receipts, technology can help translate receipt images into the necessary data elements for fuel-related metrics. This allows these teams to spend more time on how to optimize fuel use for decarbonization, using time for data insights instead of time chasing the data.

Q: What other roles do you see generative AI playing in addressing sustainability goals? Is there predictive functionality promised by this technology that can be utilized by governments and organizations?

Abbosh: Generative AI isn’t a silver bullet but if used strategically it can help advance sustainability initiatives and goals at scale. To get the most out of their sustainability initiatives, businesses need data. Without data, sustainability is not actionable, and it’s not operational. More business leaders are beginning to see how AI can help them achieve their sustainability objectives. Sixty-four percent of executives surveyed agree that generative AI will be important for their sustainability efforts. Generative AI can help accelerate data-driven decision making enabling businesses to drive forward sustainability in meaningful ways including optimizing production levels, minimizing waste, promoting efficient energy consumption, and meeting regulatory and voluntary reporting requirements.

Shim: Gen AI has the power to combine operational and environmental data to predict and mitigate disruptions to business operations and beyond. For example, we can use data from energy grids, weather patterns, and usage trends to predict and adjust energy distribution in real time. This helps organizations limit their carbon footprint while also boosting the bottom line with cost savings. This is applicable for any organization — big or small companies, nonprofit organizations, and of course, governments. For example, IBM and the United Nations Development Programme just launched an interactive model that uses AI to forecast energy access through 2030. The model is available for 102 countries across Africa, Latin America, Asia Pacific and the Middle East, and it will help governments to make data-driven decisions and enable users to analyze complex energy issues.

Q: How do you consider and account for the additional environmental costs associated with AI?

Abbosh: While AI and generative AI are impressive technologies that can certainly bring benefits to both business and sustainability, it’s essential to acknowledge that no path forward is without its challenges. There are many ways in which to architect an end to end AI enabled solution that minimize energy consumption and thus environmental impact. 

Shim: Companies need to be smart about their approach to AI. One of our recent studies showed that 63% of organizations will apply generative AI in their IT initiatives by the end of 2024, but only 23% of organizations integrate sustainability assessments when they first design their IT projects. And this is a problem. If we focus on the sustainability of AI from the outset, we’re not only reducing our environmental footprint but also creating more efficient and cost-effective solutions. There are several ways to reap the benefits of AI while minimizing its environmental impact.

Abbosh: Companies can also help lower energy use by choosing a hybrid cloud approach. A quarter of businesses are already leveraging hybrid cloud solutions to significantly boost the sustainability and energy efficiency of their IT operations. Nearly half report a substantial positive impact on their overall IT sustainability.

Shim: Another aspect to consider is the size of your models. Large models come with more data storage and computing costs and also require more frequent updates, which means that larger is not always better. Businesses should use the right size foundation model to accomplish what they need, which helps with sustainability as well as cost and speed. For example, IBM Granite models, trained on specific, relevant data, perform just as well as larger models. Another aspect to consider is the processing location. As Oday mentioned, taking a “hybrid cloud” approach can give you the flexibility to locate your processing close to clean energy sources. Co-locating data next to processing can also result in real energy savings over time.

Our studies also show that organizations with deeply integrated IT sustainability practices not only achieve better environmental results, but also see significant benefits in operational efficiency. For example, we have developed AI chips that are 14 times more energy efficient than previous models. This drastically reduces the power required to run AI workloads, allowing us to expand our AI capabilities while minimizing energy consumption.

Q: In what ways can the impact of sustainability efforts for businesses be quantified?

Abbosh: Despite the growing emphasis on sustainability initiatives, a persistent myth lingers: that sustainability and profitability are mutually exclusive.

Shim: We know that what’s good for the planet can be good for business. Our studies have shown that organizations that embed sustainability throughout their operations see better sustainability and financial outcomes, are more likely to outperform their peers in profitability, and have higher rates of revenue growth. Also, organizations that take an enterprise-wide approach to sustainable IT report seeing benefits in their operational efficiency.

Abbosh: The reality is that financial and sustainable outcomes can not only coexist, but they reinforce each other. Fifty-two percent of organizations that embed sustainability are more likely to outperform their peers on profitability, sales growth, and talent retention

Shim: In 2023, IBM implemented energy conservation projects across more than 130 locations globally. Through these projects, we have avoided an estimated 95,000 MWh of energy consumption and 33,000 mtCO2e emissions, while saving approximately $11 million. There are other areas, too, that go beyond cutting costs. For example, organizations that embed sustainability are 56% more likely to outperform on talent attraction.

Abbosh: We’ve seen the power of embedded sustainability first hand. Downer tapped IBM Consulting and used Maximo and Watson AIOps to harness near real-time data from 200+ trains across Australia. The analytics support predictive maintenance, reduce malfunctions, and increased train reliability by 51%. Similarly, Hera Spa tapped IBM Consulting to use AI camera vision to reduce landfill waste and enhance efficiency across 89 facilities, directing more materials towards reuse.

Shim: Sustainability is a long-term strategy. Organizations should not treat sustainability as a special case. When organizations truly bake sustainability into operations—rather than treat it as an add-on, the benefits can go as far as becoming client zero of sustainability solutions, taking learnings back to clients and helping them implement and operationalize sustainability.