While major tech companies have begun sharing data on the carbon footprint of data centers, smaller organizations and entire value chains still lack reliable methods to measure AI's environmental impact comprehensively. Experts emphasize a 'product approach' and multi-stakeholder cooperation as crucial steps forward.
- 75% of sustainability teams lack methods to track AI's environmental impact.
- Full value chain measurements need absolute consumption plus efficiency metrics.
- Collaboration and standardized data exchange vital to advance AI footprint transparency.
What happened
Organizations worldwide are grappling with how to accurately measure the environmental impact of artificial intelligence throughout their entire value chains. While leading companies such as Google and Microsoft have started to disclose carbon footprint information for their data centers, this transparency has yet to extend effectively to smaller companies and suppliers.
Reports indicate that roughly 75% of sustainability teams have little knowledge on how to capture AI-related emissions realistically, particularly in more complex categories such as Scope 3 emissions, which include indirect emissions from supply chains and business travel. This gap in knowledge and data availability poses a major hurdle for many enterprises aiming for credible sustainability reporting.
Why it matters
Accurate measurement of AI’s environmental footprint is essential for companies aiming to reduce their carbon and resource impacts. Experts emphasize that sustainability reporting should extend beyond just carbon emissions to include water usage, mineral consumption, and other resource impacts. Current efficiency metrics such as Water Usage Effectiveness can mask overall environmental loads if used alone.
Without reliable, comprehensive data, companies risk producing misleading sustainability claims and missing opportunities to decarbonize effectively across their operations and supply chains. Since larger firms must report on broader value chain emissions, small and medium-sized enterprises also face indirect pressure to provide accurate data to their customers, reinforcing the systemic importance of better measurement.
What to watch next
Stakeholder collaboration and data standardization will be central to progress. Initiatives like the Partnership for Carbon Transparency (PACT) seek to develop uniform frameworks for calculating and sharing carbon footprint data across product lifecycles, enabling enterprises to benchmark and reduce their AI-related emissions more reliably.
Meanwhile, IT asset managers and end users can help by demanding detailed data from vendors, including hardware power consumption, data center locations, and asset lifecycle information. This transparency will be critical as companies move beyond isolated carbon accounting toward integrating energy, water, and material resource metrics into their sustainability strategies.