How Companies Can Manage AI Use Through Materiality, Measurement & Reporting
As the use of AI increases, the governance of its environmental implications increasingly depends on embedding AI into materiality assessments, measurement practices, and reporting systems.
Treat AI Use as a Material Sustainability Driver
Bringing AI explicitly into financial materiality and impact assessments allows companies to see where AI changes the scale or severity of existing issues or introduces new risks or opportunities.
Map, Measure, and Baseline AI Demand
To make AI governable, companies should create an inventory of how often AI is used and establish utilization metrics over time. This enables organizations to identify growth, redundancy, and hotspots in AI usage.
Control AI Impact Through Policy and Oversight
Setting rules for appropriate AI use and establishing triggers for extra review before scaling AI are crucial. This management approach applies whether AI is developed in-house or provided by vendors.
Integration into Sustainability Systems
AI is not only changing how companies operate but also demanding changes in how sustainability systems are designed and governed. Much focus has been placed on the environmental impact of AI’s energy use, water consumption, and supply chain challenges. However, it is also essential to examine how AI is applied within organizations.
Understanding where AI is applied, how often it is used, and whether those applications are necessary is critical. This understanding sets a clear path for AI deployment, with systems in place to address any environmental and social impacts before they become problems.
Value Creation Through Responsible AI Use
Organizational leaders need to focus beyond the footprint of AI by mapping its use, defining control and review processes, developing systems for ongoing quantification, and reporting transparently. The ultimate goal is to manage AI’s impact from the inside out, ensuring that the benefits outweigh the risks while maintaining sustainability as a priority.
Bringing AI into Materiality and Impact Assessments
Financial materiality and impact assessments provide a structured process for governing AI by identifying and prioritizing significant impacts. Many sustainability topics influenced by AI use—such as energy demand, emissions, water use, and workforce effects—are already assessed in existing materiality exercises. However, an explicit examination of how AI alters the drivers of those impacts is often missing.
The International Sustainability Standards Board’s IFRS materiality guidance emphasizes financial materiality, defined by whether a topic could influence the decisions of investors or other users of financial statements. How AI is utilized within companies undoubtedly influences the risks and opportunities they face and can affect their financial position.
Assessing AI’s Materiality
Determining the materiality of AI hinges on understanding its scale and concentration—such as the situations it is used in and its embedding in critical workflows. Mapping AI use across various applications can help identify where it meaningfully alters environmental, social, or financial exposure.
Governance Through Policy
Once a basis for AI’s materiality is established, the next step is to shift towards control through policy, supported by proportional measurement of demand. As access to AI expands, it can become a default tool for routine tasks, increasing demand through duplication without sufficient oversight. Policies can set expectations for appropriate application and conditions for assessing task value.
Quantifying AI Impact
Quantification makes AI use visible over time and tracks its impact. For most organizations, measuring AI impact starts with obtaining a consistent view of utilization and its evolution. This foundational understanding supports precise attribution of energy or emissions and helps establish a baseline for effective identification of growth and overall impact.
Managing AI’s Impact
For organizations that own or operate their AI infrastructure, management responsibility lies within established operational controls, including decarbonization of electricity supply and hardware lifecycle management. Governance should also cover model training and retraining, especially in areas with concentrated energy and water demand.
For AI capabilities accessed through third-party providers, these impact areas must be addressed through policy and supplier engagement practices that link disclosure with procurement decision-making.
Conclusion
AI’s sustainability effects depend on infrastructure efficiency, energy sources, and governance of its use within organizations. Effective management includes assessing material impacts, setting policies for demand monitoring, measuring results, and ensuring transparent reporting. Treating AI as a source of managed sustainability can help mitigate risks and ensure that the environmental and social effects of AI use align with value creation.
