Integrating AI into ESG Risk Scoring

ESG Risk Management /Sustainable Investing

(Source: Barnes & Thornburg)

Amid increasing urgency to address climate change and socioeconomic inequality, Environmental, Social and Governance (ESG) risk scoring has become a cornerstone of company performance evaluation. Artificial intelligence (AI) is revolutionising this process by offering remarkable precision and efficiency in assessing ESG factors. With the capacity to process vast amounts of data across diverse metrics, AI enables comprehensive, data-driven ESG performance assessments and decision-making. However, AI has also come under scrutiny due to bias and high energy consumption, raising concerns about its effectiveness in ESG analysis. Drawing upon Clarity AI and MSCI, this commentary evaluates the benefits and challenges of leveraging AI-powered platforms in ESG risk scoring.

Clarity AI and MSCI are two prominent players in ESG risk scoring with distinct methodologies and frameworks. Founded in 2017, Clarity AI has gained recognition for offering ESG data with considerable depth and coverage [1]. Its highly modular and scalable platform enables clients to access and integrate its ESG datasets into their existing workflows[2]. The customisable interface allows clients to incorporate ESG metrics using APIs, widgets and standalone applications[3]. Furthermore, Clarity AI has been a pioneer in technological innovation, using natural language processing, machine learning classifiers, and large language models to enhance data processing and analysis. In January 2021, BlackRock announced a minority investment in Clarity AI, signifying a strategic partnership to bolster sustainable investing analytics within its Aladdin platform[4].

MSCI, founded in 1969, has a long history of providing ESG indices, multi-asset portfolio analytics and climate data to the global investment community[5]. Known for its comprehensive data coverage, MSCI has developed over 900 metrics including data on carbon emissions, fossil fuel exposure and climate-related value at risk[6]. Moreover, MSCI has partnered with technology companies like Google Cloud’s Vertex AI to enhance its data analytics capabilities, supporting investors in long term risk management and climate research[7]. Hence, MSCI is an invaluable tool for institutions seeking broad asset class coverage with a focus on climate and net zero goals[8].

Comparing different AI-powered tools and platforms is valuable for sustainability-minded investors, as it highlights distinct yet often complementary strategies for evaluating ESG factors. Clarity AI’s flexible, modular platform with high transparency appeals to clients who prioritise customisation, while MSCI’s extensive dataset is ideal for institutions focused on large scale, long term risk assessment.

Despite the advances AI has made in ESG risk scoring, it has not been free of criticism. AI requires significant computational power, raising environmental concerns because of its high energy consumption. Additionally, the ESG risk scoring provided by different agencies can vary widely, potentially leading to inconsistent assessments[9]. AI models may also introduce representation bias, especially because many datasets mainly represent developed markets.

This overlooks regional differences and neglects the principle of ‘Common but Differentiated Responsibilities and Respective Capabilities’, which acknowledges that developed and developing countries face unique environmental challenges[10]. For example, an overreliance on AI-generated ratings could drive investors to withdraw from high-emission emerging markets, reducing essential funding for these regions to transition to low-carbon economies[11].

Mitigating these issues is critical to ensure that AI-driven ESG risk scoring is comprehensive and reliable. To tackle AI's high energy demands, companies are exploring solutions such as optimising algorithms for greater energy efficiency and adopting renewable energy sources. Moreover, a hybrid approach that combines qualitative and quantitative ESG analysis is essential to improve the consistency of risk scoring. A qualitative approach, for example, can involve assessing a company’s adherence to widely accepted reporting frameworks, such as the Sustainability Accounting Standards Board[12]. Additionally, incorporating diverse datasets that consider regional variations in sustainability efforts will be useful in increasing the representation of different markets. Through such measures, investors not only improve the reliability of their ESG risk scoring, but also promote sustainable economic growth. Looking forward, leveraging the vast potential of AI creates exciting opportunities to foster innovation and collaboration when driving sustainable investments.

AI-Powered ESG Risk Scoring Platforms:

Clarity AI: A newer platform known for its modular, customisable interface and integration of machine learning technology

MSCI: An established player with comprehensive ESG metrics, datasets and multi-asset portfolio analytics

Research Analyst: Clara Hong Research Editor: Ryn Tan

References

[1] ‌Clarity AI, “Use Cases for Sustainability & ESG Reporting, Assessment & Analytics | Clarity AI,” (AI-Powered Sustainability Platform 2021).

[2][3][5][6][7][8] Forrester Reprint, “The Forrester Wave™: ESG Data And Analytics Providers” (Forrester.com. 2024).

[4] BlackRock, “BlackRock Announced Minority Investment in Clarity AI,” (BlackRock. 2022).

[9] Richard Hardyment, “AI could supercharge ESG, but only if people remain in charge,” (Thomson Reuters Ethical Corporation Magazine 2024).

[10][11] IBM, “AI Bias,” (Ibm.com, 2023).

[12] SASB, “SASB - ESG Reporting - Getting Started with SASB,” (SASB, 2024).‌