Glenda Korporaal, link here
British consultancy GaiaLens is using artificial intelligence to step up its screening program to combat greenwashing among Australian companies.
The firm, which has just opened its Australian office, scans global media and other public data sources to see if companies are meeting their environmental, social and governance (ESG) commitments, supplying its data to major investors.
The company’s founder, Gordon Tveito-Duncan, who recently visited Australia, said there was increasing interest among superannuation funds and institutional investors in Australia about the green credentials of the companies they invested in.
The AI-powered GaiaLens platform has already analysed 700 leading Australian companies with regard to their ESG credentials.
Mr Tveito-Duncan said it planned to expand the database of sources it analysed, including whether a company received a fine or a penalty from a regulatory agency.
“As the world transitions from fossil to renewable forms of energy, and as investors increasingly seek to adhere to environmental, social and governance principles, the most immediate interest and area of focus from institutional investors is around the issue of greenwashing,” he said.
“No matter where people stand on ESG issues, the concern around greenwashing is extremely real as investment houses seek to build trust and confidence, protect their fund inflows, and avoid regulatory scrutiny and shareholder activism.”
Mr Tveito-Duncan, and his friend Seb Kirk, became interested in the potential for a consultancy around ESG in 2019 when they were studying a master’s course in data science at the City University of London.
But it was their ability to leverage the tools of artificial intelligence which allowed them to develop their own screening tool for thousands of companies, launching their company in 2021.
“GaiaLens uses proprietary machine learning (ML) models to engineer ESG factors which are then inputed into its ESG scores algorithm,” Mr Tveito-Duncan said. “It also uses ML models to calculate forward-looking ESG scores.”
He said the platform used the latest large language model (LLM) technologies to process the news at scale.
“We have fine-tuned an LLM on ESG news data, and for each article that it processes, it calculates four controversy scores (E, S, G and a greenwashing score) and over 50 ESG sentiment scores,” he said.
“The real game-changer with this approach is that the model provides reasoning for each score, explaining why the model assigned a particular score to an article,” he said.
“This kind of output was not possible until recently.”
The firm uses AI to study data for some 20,000 globally listed companies and two million private companies, searching data sets for equities, bonds, infrastructure and global markets.
The two founders visited Australia in 2023 as part of a trade mission.