Revolutionizing Climate Reporting with Generative AI

In response to the changing climate and its impacts on the global economy, pressure from investors, regulators, and consumers to increase transparency has led to regulations that increasingly require businesses to evaluate and disclose significant climate risks and opportunities facing their operations. Such disclosure and supporting analysis encourage companies towards responsible business practices that promote climate resilience and facilitate a well-managed economic transition towards a more low-carbon future.

Voluntary and mandatory climate reporting is a rapidly evolving space. Global climate disclosure regulations such as the US SEC’s recently adopted climate-related disclosure rules, the EU’s Corporate Sustainability Reporting Disclosure (CSRD), and voluntary frameworks such as CDP (formerly known as the Carbon Disclosure Project), vary in scope, detail and timelines for compliance. Though there are some commonalities, these disclosure requirements differ in their level of prescription and exhibit varying degrees of standardization. This can lead to challenges in adoption by reporting entities and in interpretation by investors, consumers, and regulators who are becoming more reliant on holistic and consistent climate disclosure. To address this gap, generative AI (GenAI) can help assess the quality of existing climate disclosures and potentially generate more thorough disclosures based on the analysis of market responses in conjunction with an AI-driven analysis of company climate data.

In this talk, we will present how Gen AI can be used to evaluate climate disclosures, identify gaps against human-identified industry benchmarks, compare against peers within and across industries, and provide actionable recommendations for improvement. For example, a company can use Gen AI to run a climate scenario analysis and evaluate the results such as identifying drivers of a forecast of increased tropical cyclone severity for a region of the world. We will show how Gen AI has the potential to uncover trends or causalities beyond the surface level of such a climate scenario analysis, a type of analysis that would take considerable human time and effort. However, it’s important to note that human oversight remains crucial in guiding AI-driven insights, helping accuracy, and providing context. Finally, while we will demonstrate how AI can generate thorough and robust climate disclosures by leveraging insights from both leading practices and company-specific climate analyses, we will also emphasize the importance of responsible AI practices, including human-in-the-loop controls, to address the limitations of Gen AI. These safeguards are essential to help AI-generated disclosures meet the stringent standards of regulatory scrutiny, such as those required by the SEC and other governing bodies.

AI-driven assessments have the potential to improve the quality of climate disclosures, making them more transparent and actionable. This approach not only helps companies meet regulatory requirements but also enhances their preparedness for climate-related risks, ultimately promoting better risk management and sustainability, and provides external parties with a more complete understanding of the climate risks facing the organization.