What it means
AI systems, particularly large language models and generative AI, consume significant computational energy and water resources. As ESG reporting obligations expand and regulatory scrutiny of AI’s environmental impact increases, organisations without visibility into their AI’s resource consumption face disclosure and reputational exposure.
Why it matters
Boards with ESG commitments or sustainability reporting obligations cannot exclude AI from their environmental accounting. Regulators and investors are beginning to require transparency on AI-related resource consumption. An organisation that cannot account for its AI’s environmental footprint is not meeting its disclosure obligations.
Board governance implications
The board must confirm that AI procurement decisions include assessment of energy and water consumption, that this data is incorporated into ESG reporting where required, and that vendor environmental claims are subject to due diligence.
Governance failure timeline
Pre-deployment
Failure to include energy and water consumption assessment in AI procurement decisions, and to confirm that AI-related resource data is incorporated into ESG reporting obligations, before tools are approved for organisational use.
Deployment
AI operations are consuming energy and water resources that are unaccounted for, inconsistent with stated ESG commitments, and absent from sustainability reporting while the systems are live.
The gap between what the organisation says and what it does is accumulating, invisibly.
Post-deployment
Investors and regulators identify the reporting gap.
Where AI resource consumption contradicts sustainability commitments, the reputational exposure is significant.
Where disclosure is mandated and absent, a regulatory breach follows.