As AI becomes part of everyday business operations, companies are increasingly being asked: what’s the environmental cost? Whether you’re working toward ESG goals or preparing for climate disclosures, measuring AI’s impact is essential.
What to measure:
- Carbon emissions from usage (Scope 3 if using 3rd party tools like ChatGPT, Claude, Gemini, and Perplexity)
- Water use from data center cooling
- Energy consumption if operating models in-house
Methods:
- Spend-based approach: Estimate impact based on software spend and known emissions factors.
- Usage-based approach: Track number of queries or compute hours used.
- Browser extensions/tools: Some tools help track AI use across platforms.
Why it matters:
- Meet climate disclosure standards (SBTi, CDP, SEC rules)
- Inform procurement decisions
- Identify opportunities to reduce usage or switch to lower-impact providers