The Risks of Artificial Intelligence: How Investors Can Deploy Sustainable Technology
The risks associated with artificial intelligence (AI) often make headlines, but how can the investment community harness this technology for sustainable purposes?
As investment professionals integrate environmental, social, and governance (ESG) factors into their decision-making process, they face a significant challenge: the quality of these data.
Rapid Progress
This is a relatively new field in finance, and the data can be fragmented, outdated, unverified, or even non-existent. They can be inconsistent and difficult to compare. Certain ESG factors, such as social impact, are challenging to measure. Investors may be overwhelmed by mountains of information and unsure of what is important or how to apply it. While the availability of ESG data has increased in recent years, not all of it is reliable. The use of flawed ESG data exposes investors to financial and reputational risks. Furthermore, regulatory scrutiny of fund labeling, particularly in developed markets, has increased the risks associated with using mediocre or insufficient data.
The progress of AI has fueled hope that this technology can help bridge the gaps in ESG data sets, just as it aids investors in performing more general analyses. For example, AI played a crucial role in developing the ESG taxonomy of the Sustainable Development Investment Asset Owner Platform, whose member asset owners manage $1.5 trillion in assets. This initiative is designed as an investment standard aligned with the United Nations' Sustainable Development Goals (SDGs) and aims to measure the contribution of assets, such as companies, to these SDGs.
The Power of Processing Capacity
AI can contribute to improving the availability and reliability of ESG data in several ways. Many asset managers already use AI, directly or through data providers, to find and process data that they can use to identify sustainable investment opportunities, emerging trends, or imminent risks to their portfolios. Some investors use data filtering services to exclude investments that do not meet their ESG criteria. The ESG data that investors need are likely scattered across numerous sources and formats. AI can be used to scour hundreds of thousands of publicly accessible sources in multiple languages. This includes company websites, sustainability reports, news articles, press releases, independent research, conference call transcripts, and increasingly, social media. In brief, the technology can help investors access more information than human analysis would allow, and do so much faster. Natural Language Processing (NLP), a branch of AI in which computers analyze language like humans, can detect sentiment in text. If customer complaints about a company multiply or a controversy begins to brew on social media, the technology could identify an imminent ESG risk before a human analyst does or before the company's stock price is affected. NLP can signal information about companies that pollute or mistreat their employees before it gains traction. It can also quantify difficult-to-measure factors, such as employee satisfaction, which could be a useful measure of a company's social performance.
Boost to Reporting
AI can also monitor how corporate activities affect biodiversity and ecosystems, for instance, by determining whether a company contributes to deforestation or produces waste or air pollution. It can even be used to sift through satellite images to detect methane emissions or environmental pollution. This could enable the identification of risks throughout a company's value chain, beyond its direct activities. Conversely, AI could discern the impact of natural disasters or extreme weather conditions on a company's assets and activities. AI can even help verify companies' compliance with growing reporting requirements, such as those imposed by the EU's Non-Financial Reporting Directive or the Task Force on Climate-related Financial Disclosures. Using technology to exploit data presents obvious advantages, including its potential to surpass human capabilities, as human analysis is subjective and prone to errors. But AI can evaluate vast amounts of data much faster and more accurately than humans. This could enable asset managers to integrate more ESG factors into their investment decisions.
Security Concerns
However, there are limitations. Analysis based on publicly available data published by a company is only as good as the data itself. Although reporting requirements are increasing, ESG information is not yet standardized and only covers large companies and certain markets. Executives may avoid using negative language to evade AI-powered sentiment analysis of company reports. The proliferation of data means analysts must make sense of different data sets. Furthermore, while using AI to exploit ESG data presents advantages, it also carries environmental and social risks. AI requires significant computing power. The global AI energy consumption could exceed that of many small countries, significantly increasing carbon emissions. Some experts also warn about the harm that technology could cause to society. The most tangible and immediate fear is that AI could lead to job losses, or worse. According to some estimates, the automation resulting from generative AI could impact 300 million jobs. AI could represent an existential risk to humanity if it were to start doing things that humans do not want it to do. There is also concern that AI could perpetuate biases because it is based on training data that may include human biases. Generative AI would pose a serious threat due to its ability to manipulate and deceive populations. Similarly, AI that exploits personal data or tracks individuals' online activities raises privacy concerns. In the US and Europe, legislators and regulators have expressed concerns about AI use in financial services, particularly regarding confidentiality and cybersecurity risks. Asset managers will need to consider the possibility that new regulations on AI may be introduced in the coming years before investing in the development of the technology. In any case, AI will significantly aid productivity and represent a force for good. Asset management is a field where this technology has already had a positive impact, serving as an effective tool for exploiting ESG data.
