While many corporates are turning to AI to help manage sustainability data and reporting, other important stakeholders are starting to use it for other things that may not be so positive for companies. Responsible Investor reports that
“some investors are turning to ChatGPT and other generative AI assistants as a cost-effective tool to monitor ESG controversies. Norges Bank Investment Management (NBIM), which manages Norway’s $2 trillion sovereign wealth fund, has been using Claude to monitor controversies since May, according to Cristopher Wright, head of ESG risk monitoring.
‘AI is particularly powerful for assessing ESG risk because so much of the underlying information is inherently qualitative,’ he told Responsible Investor…
For ESG controversies, tests subsequently showed that the AI tool ‘was capable of identifying information of higher quality than what we can buy from external data vendors’ and doing it at lower cost, Wright said. Claude is now being used to monitor in real time the roughly 8,500 companies in NBIM’s portfolio, something the fund was not previously able to do…
Other investors are using AI assistants to keep tabs on controversies in areas where there are limited or no alternatives on the market, such as human rights risks in conflict zones. Ulla Benediktson, head of sustainable investments at Velliv, said the Danish pension fund uses AI ‘primarily to address gaps in coverage from established data providers, for example limited coverage of themes such as exposure to conflict areas’.”
This approach may be quick, easy and cost-effective, but there are reasons to be concerned about validity and accuracy of the results. Data governance and interpretation with embedded algorithmic bias are issues already under scrutiny broadly in AI. I’ve written many times about my disappointing experiences with AI in sustainability (here and here, for example). Other times, I asked AI to design specific physical mechanisms I wanted to build. In one instance, it couldn’t differentiate between a bipod and a tripod, and in another the simple mechanism violated basic rules of science in two separate ways. It worries me that unchecked, unvalidated AI output is being relied on to identify ESG controversies. You should be concerned, too.
Members can learn more about AI and sustainability here.
Interested in a full membership with access to the complete range of benefits and resources? Sign up now and take advantage of our no-risk “100-Day Promise” – during the first 100 days as an activated member, you may cancel for any reason and receive a full refund. But it will probably pay for itself before then. Members also save hours of research and reading time each week by using our filtered and curated library of ESG/sustainability resources covering over 100 sustainability subject areas – updated daily with practical and credible information compiled without the use of AI.
Practical Guidance for Companies, Curated for Clarity.
