• Josh Hasdell, Author |
5 min read

Before joining KPMG, I trained and worked as a marine scientist, learning about how the largest animals on the planet interact with some of the smallest microorganisms and how they’re all inextricably interconnected to sustain life in the ocean and on the planet. Speaking at a KPMG event last winter in Vancouver on “AI for Good,” one of my key messages was that taking the view only from the shoreline can’t give a full picture of the complex systems at work beneath the surface. Understanding how to track and compute all this information and turn it into something actionable for conservation is therefore a critical challenge.

This same analogy can be applied to the business world. Organizations looking to develop an integrated environmental, social and governance (ESG) strategy need to look beyond their own operations and consider ESG issues across industries and sectors, as well as up and down the value chain—from manufacturing to transportation to consumption. They must also be aware of the interdependencies that exist between a spectrum of E, S, and G factors.

And just like trying to understand what happens deep below the ocean’s surface, it’s often difficult to see those interdependencies in a vast sea of data.

That’s where AI is increasingly showing promise. It isn’t just about task automation, chatbots, personalized shopping and self-driving vehicles. AI is also being applied to many of the challenges we’re facing in the ESG space.

A data-driven approach
According to the World Economic Forum (WEF), climate change is poised to become a major business disruptor over the coming decades. “To adapt to this new reality," the WEF says, “more granular climate insights need to be generated to empower stakeholders to take a more data-driven approach to climate adaptation.”

The WEF recommends using AI to solve for a variety of challenges related to interpreting and analyzing ESG data. For example, AI can be used for:

  • Predictive modelling of extreme weather events
  • Developing early-warning systems, which could even help with pre-emptive humanitarian efforts
  • Building climate resilience, such as pinpointing potential vulnerabilities in the value chain
  • Ensuring ethical supply chains and protecting human rights and social value.

To be sure, historical notions of “value” are shifting as capital markets take note of whether a company is acquiring profit off the backs of racialized, marginalized or vulnerable populations. AI can lead to better social and corporate governance impacts or outcomes through multi-dimensional insights.

Emerging capabilities
Animals and ecosystems have spent millions of years finding the most efficient and effective ways of living and thriving together in nature. AI does this too—only much faster.

AI can digest massive amounts of data for detail-oriented analysis, while machine learning identifies patterns in data sets and learns optimization routes without being explicitly programmed to do so. Deep learning, a subset of machine learning, mimics neural networks in the human brain, allowing for highly accurate predictions.

[Related: Feite Kraay: “You’ve come a long way, AI!”]

Many business leaders are turning to AI for climate mitigation, harnessing its power to perform complex tasks such as measuring and reducing carbon emissions, or optimizing energy usage by balancing peaks and valleys on the power grid. But it can also be used more broadly, to acquire insights and enable better decision-making.

For example, temperature data going back to the 1850s can help us see how temperatures have risen over time. But understanding climate change—to both mitigate it and adapt to it—requires understanding more than a chronological set of temperature ranges. Using AI, it’s possible to start layering on other factors to understand broader related societal and geopolitical risks.

The challenge is to improve data quality to obtain more accurate disclosures, advance AI objectives and address growing compliance requirements—and to derive better insights. In the shipping industry, for instance, KPMG is working with a maritime solutions company to incorporate AI into sustainability initiatives such as fuel and energy management. By creating a model to understand the correlation of data sets, they’re able to better understand shipping and traffic impacts across the value chain. Reducing the number of valves, for example, allows them to reduce the energy spent on sailing and, in turn, reduce their carbon impact.

Gaining deeper insights
Many business leaders have only been taking a one-dimensional view of ESG: how it affects their enterprise value. But AI is also transforming ESG through its ability to manage multiple data sets, in multiple formats, across various sources and in various states of quality—which isn’t otherwise possible using today’s technology stack without AI.

That’s because AI can provide a three-dimensional view, helping business leaders understand the connections and interdependencies between their operations and ESG strategies:

  • A 1D view looks at how ESG factors are affecting a given company (e.g., single materiality)
  • A 2D view looks at how those factors impact the company’s enterprise value within society and throughout the value chain (e.g., double materiality)
  • And a 3D view adds in dynamic assessments that include velocity and severity (e.g., dynamic materiality).

On the output side, AI can help to find efficiencies that will reduce CO2 emissions and help organizations meet established targets, as well as enable ecosystem and value chain integration to reduce “Scope 3” emissions.

By modelling climate resilience with digital twins, for example, it’s possible to test what/if scenarios, forecast hazards and futureproof operations—ultimately helping to protect people and communities. Companies can even use digital twins to test circular economy models by creating simulations to validate those models before breaking new ground.

AI could also be used to help track and report on ESG initiatives. For example, it could accurately report on the success of a recycling program by using “computer vision” to determine which types of bottles and how many have been recycled in a particular region compared to how many were sold in that region. This will become increasingly important as companies face more intense scrutiny over greenwashing.

By moving from 1D to 3D insights, business leaders will have a better understanding of the labour and social impacts of decarbonization. This could help them hold third- and fourth-party suppliers accountable for human rights issues, or mitigate negative impacts on workers, local communities, Indigenous groups and the environment.

Taking the plunge
Getting started with AI in ESG begins with understanding what role AI can have in transforming your business and under what circumstances. But humans and AI will need to work together. “We always need human creativity and judgment in those systems,” says Aya Ladki, a manager at KPMG Ignition Vancouver who also spoke at the “AI for Good” event. “It’s the equitable input of both human and technology that will secure the right balance and the right approach.”

Note, too, that a variety of stakeholders should be involved in these conversations, not just the IT team. When all stakeholders can sit at the same table and agree on a sustainable framework for AI, they can then start to bring into focus the complexity beneath the surface—and solve some of their greatest ESG challenges.

Multilingual post

This post is also available in the following languages

Stay up to date with what matters to you

Gain access to personalized content based on your interests by signing up today