The simultaneous rise of artificial intelligence and environmental, social, and governance in the business sector creates an opportunity to use ESG to better regulate AI. Including AI usage and policies in ESG reports could bring needed clarity for consumers, investors, and public officials.
ChatGPT was launched by OpenAI in November 2022 as an AI driven chatbot. ChatGPT became instantly popular, with some in the business sector utilizing it to do a variety of business functions. However, its use is not without problems. Its abilities are limited, and the work produced has been questionable at times.
OpenAI is also facing legal issues over the safety of their product. The FTC is currently engaged in an investigation see if the company “engaged in unfair or deceptive practices relating to risk of harm to consumers, including reputational harm.” As part of the same investigation, the FTC is also looking at OpenAI’s data and privacy standards, suggesting that users’ confidential information is being shared.
In a broader context, use of AI raises legal and ethical concerns. AI can simulate likenesses of individuals, possibly usurping existing image and likeness laws. This debate has become part of the writers’ strike by SAG-AFTRA. AI generated images and deepfake videos are creating confusion on social media, and it will only get worse as the technology improves. There is also an unsettled question as to who owns the copyright for an AI created piece of art or text. As is common in fast moving advances, the technology is outpacing the existing laws.
ESG has faced the same problems as it grown in usage over the past few years. ESG is a form of investing where non-financial actions of a company are considered. It is a tool for investors to look beyond the profits and losses of a company to other factors when making investment decisions. As the name indicates, ESG is divided into three categories addressing the environmental, social, and governance policies of a company.
Over the past three years, companies have begun producing annual ESG reports highlighting their ESG related actions and policies. Until the past few months, reporting standards for ESG were nonexistent. Only July 31, the European Union adopted the European Sustainability Reporting Standards, which created their framework for ESG reporting. A month earlier, the International Sustainability Standards Board announced the International Financial Reporting Standards Foundation Sustainability Disclosure Standards.
The IFRS Standards are focused on how companies report their carbon emissions and environmental policies. The ESRS will incorporate the IFRS Standards for environmental reporting. Likewise, the U.S. Securities and Exchange Commission is set to announce their own disclosure standards focused exclusively on sustainability and environmental policies.
The social and governance categories have received significantly less regulatory attention in both the EU and the US. However, it is in the governance category where ESG may help in AI regulation.
As the usage of AI increases in the corporate world, so has the demand for AI policies. The policies may include data privacy and security, confidentiality, and ethics. While they are not required at this time, regulatory development and consumer demand may soon make them a common and required practice. Investors may also wish to take note as to a company’s AI policy, as questionable and unethical usage of AI may adversely impact a company’s long-term profits.
What is reported in governance category of ESG is largely undefined. In generally refers to the company’s internal policies relating to ethics, compliance, and oversight. Some have interpreted that to include data management and cyber security. I propose that is also a logical fit for disclosure of a company’s AI usage and policies.
The choice to include AI policies in ESG reports is based solely on the priorities of the company. If a company feels inclusion will be to their benefit, we may start seeing it in the near future. However, the demand for regulatory development in this area will only increase. AI policies will soon be required, how they will be shared with the public is still to be decided.