AI Knows if You Are Guilty of Greenwashing

Talk is cheap regarding companies talking up their credentials in Environment Social and Governance (ESG). But artificial intelligence and natural language processing can help identify those who are more serious about it than others.

At U.S.-based fund manager Acadian Asset Management, the firm has developed a tool that uses artificial intelligence and machine learning to rank corporates by the seriousness of their intent.

In the investment world, many fund managers base their decisions on which companies to invest in on how much they disclose about their activities. But this is only part of the story, says Acadian’s director of responsible investing, Andy Moniz.

To get the real picture requires the skills and applications of an investigative journalist. This is impossible to do at scale when assessing a universe of 40,000 investable securities based on their ESG actions and not just their words.

“We’re merging these datasets together to identify companies that talk a lot about sustainability but perhaps don’t actually do much”

“Historically, companies which disclosed information were rewarded through higher ESG ratings simply for being transparent in their reporting,” Moniz explained at a Sydney presentation this week.

“This creates a perverse situation where a fossil fuel company can receive a higher ESG rating than an electric vehicles company. It’s very easy for companies to make aspirational sustainability commitments. Although around three-quarters of companies have announced decarbonization targets, it’s a very different story as to whether those targets are actually aligned to science-based targets.”

Not scalable

While the traditional approach in assessing companies is to meet with management and ask them probing questions, this is not a scalable solution when seeking to construct a global portfolio.

Acadian’s AI solution, called Engager, uses an algorithm to examine company documents such as sustainability reports, policies, press releases, regulatory filings, and even executives' voices on analyst calls.

“We’re merging these datasets together to identify companies that talk a lot about sustainability but perhaps don’t actually do much,” said Moniz.

“We also identify evasive and potentially deceptive talk. The algorithms can assess to what extent managers are directly answering sustainability questions or giving a boilerplate response or an indirect answer.”

Boastful CEOs

Using natural language processing, the algorithm can identify when companies fail to answer questions adequately on an earnings call, give qualified answers when they talk about sustainability issues, and when there are caveats to their answers.

A CEO might boast, for example, that his company is a world leader in technology and innovation, “the envy of the rest of the world.”

“Using AI, we can identify this or 40,000 securities across different languages. The way the algorithm works is that it looks at the words in the question and looks at the words in the answer, and it says that clearly ‘this is an evasive answer,’” said Moniz.

He also gave the example of a transcript from an earnings call, where an analyst asked management about the company’s decarbonization plans and implications for capital expenditure.

The algorithm has been trained to understand what green leaders and greenwashers say and ranks the replies accordingly based on linguistic cues.

“Another example is when we look at referral statements, this is where managers are asked a question, and they don’t answer, or they say ‘oh, we answered that previously in last year’s earnings call or looked at our reports from two years ago,’’ said Moniz.

“These statements are designed to provide some credibility, but of course, extremely difficult to cross verify, and an analyst needs to identify the information and prior reports manually. And again, we can use our technique to connect the dots to identify these patterns and to hold companies to account.”

The Engage platform is just another of many examples of the contribution AI can make to sustainability goals.

According to PwC research commissioned by Microsoft, using AI for environmental applications could contribute up to USD5.2 trillion to the global economy in 2030, a 4.4% increase relative to business as usual.

In parallel, AI levers could reduce worldwide greenhouse gas (GHG) emissions by 4% in 2030, equivalent to 2.4 Gt CO2e — equivalent to the 2030 annual emissions of Australia, Canada, and Japan combined.

Weeding out the greenwashers is a good start, but it is just one of the many applications of AI that can make a positive contribution to the cause of sustainability.

Lachlan Colquhoun is the Australia and New Zealand correspondent for CDOTrends and the NextGenConnectivity editor. He remains fascinated with how businesses reinvent themselves through digital technology to solve existing issues and change their entire business models. You can reach him at [email protected].

Image credit: iStockphoto/Tanaonte