AI: A Profitable Path to Sustainability

By Philip Smith and Eduardo Plastino

Next week, we will be in Denver, Colorado, for the North American Sustainability & Responsibility Summit (NASRS). It’s an excellent opportunity to meet current and potential partners, clients and other ecosystems players, and, as always, to learn from each other and share the lessons we’re all gathering along the way.  

It’s a large, passionate and diverse community and the issues and opportunities we’re all trying to solve are wicked; complex, interconnected, rapidly changing. To a high degree, that necessitates that we collaborate well, but swim in our own lanes so our contributions are differentiated and impactful.  

So, NASRS has also made us think harder about the aspect of sustainability we want to concentrate our conversations on. In other words, consider all the many ways in which digital technology (the focus of Cognizant’s business) can further sustainable business outcomes and select a specific topic of conversation we thought it made sense for us to bring to the wider discussion at NASRS. 

Why AI  

We went for AI and will be hosting a delegate meeting entitled 'SustAInability: Exploring how AI and other digital technologies can unlock and accelerate our responses to environmental risk and opportunity'.  

Why? Well, our daily work with clients leaves no doubt that these two macro trends - sustainability and the AI/wider digital revolution - must go hand in hand if we are to address key aspects of the polycrisis our world finds itself in and ensure our businesses are tools for enduring value creation for all. Many refer to this as the triple bottom line: people, planet and profit.  

However challenging their market contexts, business leaders not only increasingly understand this, many of them also see - with greater clarity and engagement - the immense opportunity this economic and technological transformation toward sustainability offers. 

Many, yes, even most - but not all. In our recent survey of senior executives with oversight of their companies’ sustainability efforts, 58% said their firms are already using the power of AI in this area. Worryingly, this still means that over four in ten have been slower to act. 

42% of companies still don't use AI for sustainability

AI is far from the only technology that can help businesses boost their sustainability and is certainly not a panacea, but it is a highly powerful tool – so much so that three in four of those that do use it for this purpose find it effective or highly effective. Their success should serve as an inspiration to the 42% that are still missing out on a great way to add value to all elements of the triple bottom line.  

We can achieve so much more. A recent study by Google and BCG found that scaling existing, proven AI applications could generate insights that would help reduce global greenhouse gas emissions by between 5% and 10% by 2030. 

Sound business case 

Getting there and increasing its uptake and effective use, however, will depend on companies seeing the sound business case for this use of AI. A look at key aspects of business’ sustainability impact reveals just how strong this case can be.

1. Processes and operations 

We recently published a series of articles exploring how companies from various industries can move towards more sustainable business models. AI and other digital technologies play a key role. Examples include: 

• Using AI to interrogate the large quantities of data generated by sensors and smart meters to produce more accurate electricity supply and demand forecasts. These help keep the balance of smart grids in which electricity flows both ways between distributors and the growing number of consumers who are also producers, the so-called prosumers (thanks to the proliferation of domestic-scale solar and storage technology).  
• Generating data-based insights for everything from current and likely changes in soil health to short- and long-term weather forecasts to support regenerative agriculture techniques. 
• Helping financial services companies calculate the emissions they finance through credit and investments (the first step to address them). 

2. Product design 

We can use AI to better assess the recyclability of existing and innovative materials to improve waste and resource management practices. Staying with our precious natural resources, applying AI in the product design phase can reduce waste as well as costs. A case in point is that of SXD Zero Waste, which uses AI to redesign garment mock-ups. This application of the technology allows it to create pieces such as dresses, sweaters and pants with almost no fabric waste - compared with between 10% and 30% in traditional designs - at 55% lower costs.  

3. Closing the loop 

'Traditional', linear value chains start from gathering raw materials (sometimes sustainably, sometimes less so) and continue in stages such as processing or manufacturing, transporting and so on, and all too often end with the disposal of the final product once it has been used or its productive life has been depleted.   

By contrast, circular economy business models seek to ‘close the loop’, so that waste at any stage of the process can be used symbiotically as input elsewhere. In many activities, this requires identifying which materials are suitable for which processes and which stakeholders - and that can be tricky. 

Enter AI - and consider the case of ZenRobotics, the first company to combine AI and robotics in a waste processing environment. The company uses its AI software to analyze waste stream data collected by sensors. This information is used by heavy-duty robots to autonomously ‘decide’ which objects to pick and send where. The process fully automates the performance of waste sorting, making it more accurate and allowing more and higher quality materials to be recovered and used as inputs for the production of new products. 

What about AI’s own environmental impact? 

While these and many other use cases are exciting and indeed vital, we can’t ignore that AI itself can also harm the environment. The media and other commentators have covered the enormous implications of the higher levels of energy consumption necessary for intense compute activity in data centers as a result of the ‘physical labor’ of training and running AI applications – not to mention the large freshwater volumes needed to cool these environments.  

AI is essential to tackle our large environmental problems, but it also has a substantial sustainability impact - and we need to address it.

We absolutely need to address these issues - for example, by seeking to ensure that data centers are not just built in areas where abundant renewable energy is available, but also produce as much of their own clean energy as possible. Their owners also need to be fully transparent about the environmental impact of AI. Many leading players are moving in that direction; this transparency is vital to increasing levels of trust which will encourage others to follow suit. 

Moreover, all of us should also always opt for coding (including reuse of code and models) and digital tools with lower environmental impact (great ideas and resources for this are available by the Green Algorithms project and the Green Software Foundation).   

In any case, acknowledging that there is (much) work to be done in improving AI’s sustainability footprint must not lead us to throw the baby out with the bathwater. 

Looking ahead 

A study led by Ricardo Vinuesa , of the KTH Royal Institute of Technology, in Stockholm, has looked into whether AI helps or hinders humanity’s efforts to achieve the UN Sustainable Development Goals (SDGs). These are a series of 17 health, education, equality, environmental, socio-economic and growth-related objectives, each one broken down into specific targets, that countries should seek to achieve by 2030.  

In the case of the three environment-specific SDGs, AI was found to be an enabler of 25 targets but an inhibitor for just 8. The obvious conclusion: it’s clearly a net positive. We should make the most of its benefits and work to address the difficulties while also being open with and interrogating and removing those inhibitors. This way, we will be able to unlock the full potential of this remarkable tool to help humanity and the rest of nature thrive.  

All in all, there is little doubt that AI and digital technologies can be formidable allies in addressing our pressing sustainability problems and creating substantial value for businesses and all their stakeholders. We look forward to continuing to walk this road with our clients, partners and the wider ecosystem. Next stop: Denver.