Sustainability investors are looking to artificial intelligence (AI) technologies to assist them accomplish their ESG goals and financial performance while assessing potential risks.

Recent advancements in artificial intelligence, such as machine learning and natural language processing, can be powerful tools for sustainability-focused investors, as a variety of AI-powered solutions promise to assist investors in navigating companies’ financial performance prospects as well as environmental, social, and governance (ESG) considerations.

“The integration of AI into sustainable investing could mark a profound turning point in investors’ ability to navigate the complex web of ESG factors,” says Deviscorps’s Head of Global Sustainable Finance, Matthew Jordan “By harnessing AI’s analytical capabilities, investors can identify companies with strong ESG performance, mitigate risks and shape portfolios that better align with sustainability objectives.”

Investors interested in using AI in support of their sustainable investing and ESG objectives should consider the types of applications currently in market, as well as the potential risks and key questions to ask when assessing these new tools.

AI Applications in Sustainable Investing

Some investors are already adopting AI technologies for various purposes. Three of the most prominent examples include:

Predictive models to address ESG disclosure gaps: Machine learning, a sort of artificial intelligence inspired by human brain processes, has the potential to increase the accuracy of ESG measures. These measures are in great demand: While 88% of asset owners in a Morgan Stanley Institute for Sustainable Investing survey said ESG information was crucial in choosing an asset manager, just 39% of asset managers provided this reporting and transparency.1 Meanwhile, only 35% of publicly traded businesses worldwide disclose at least some of their greenhouse gas emissions—an increasingly important risk metric for investors.

Predictive modelling, which employs machine learning techniques and publicly available data, is assisting investors in filling in the gaps in sustainability reporting. Currently, when estimating greenhouse gas emissions, investors frequently utilise an industry average for companies that do not report emissions or simple linear extrapolation to model emission values based on company-disclosed characteristics. Machine-learning methods, on the other hand, can detect new parallels in data based not just on industry, but also on location, revenue breakdown, and product and service categories. Identifying these data links may lead to more accurate projections.

Natural language processing to assess sentiment and risk: Natural-language processing (NLP) enables investors to analyse hundreds of media and other sources of information on a daily basis, overcoming the limitations of human data gathering and risk assessment, such as subjectivity and limited capacity. This approach can be used to discover corporations with contentious ESG practices, such as claims of human rights violations or corruption and bribery, which companies may not reveal but could be important information for investors. NLP tools can be used to detect and gather internet comments and claims in real time, giving investors valuable information about public perception and its possible impact on corporate stock prices.

AI-powered satellite sensors can evaluate a company’s vulnerability to physical dangers or negative environmental repercussions, allowing investors to make more informed judgements. These methods can recognise patterns from a huge number of inputs, such as infrared photos, in a timely and precise manner.

Currently, ESG grading agencies utilise images of deforestation and reforestation to evaluate the quality of volunteer carbon offsets. Another powerful use, the Methane Alert and Response System (MARS), monitors methane leaks, a greenhouse gas that traps more heat in the atmosphere than carbon dioxide. MARS, which will be unveiled during the United Nations Climate Change Conference in 2022, would analyse data collected by global mapping satellites to locate and attribute concentrated areas of methane emissions to a single source. This information will be made public, allowing investors to directly interact with a company about its strategy for lowering methane emissions.

Understanding the Risks

When researching AI products, ESG investors should evaluate potential risks such as data privacy and security. Because AI models require a wide range of data, including individually identifiable and sensitive information, there are concerns that AI may be used to track private behaviour, which may become publicly available through reverse engineering.

There are also concerns about the potential lack of trustworthiness and accountability in AI-generated material. Despite the hype surrounding generative AI, information created by huge language models may not be reliable or transparent in its sourcing. Without controls assuring transparency and accountability, AI tools could be used to spread discriminatory language or even promote misinformation that damages the integrity of the global financial system. Outlining explicit accountability for model and data outcomes, as well as the ability and willingness to communicate model logic and outputs, are all necessary protection mechanisms. Bias can potentially enter the system. If the training data is not representative of the population and contains algorithmic or human biases, the output data may also be biassed.