Grove Launches Open-Source Approach To Sustainable AI Measurement

Grove Collaborative launched a new open-source approach to sustainable AI measurement via an expanded measurement partnership with Gravity as well as made new sustainability commitments while calling on the industry to adopt AI mindfully, officials said.

The formula, created by Gravity, uses compute time, server power and grid emissions factors to determine AI-related carbon emissions so that companies can understand the environmental impact of their AI adoption. A baseline understanding gives companies the information they need to explore mitigation and reduction strategies so that sustainability and innovation can advance together.

“AI is essential to Grove’s future and – like the rest of the industry – we’re testing and learning with tools as they become available to stay competitive, be more efficient, and fuel growth,” said Jeff Yurcisin, CEO of Grove Collaborative. “While we’re still learning about the climate impacts of society’s AI adoption, we will not and cannot choose between innovation and sustainability. Being a sustainable business and adopting artificial intelligence are not at odds, but every organization must prioritize ethics, emissions tracking and environmental stewardship in their AI strategies – regardless of the size of that impact. We estimate our 2025 AI-related emissions will equate to roughly 6% of last year’s business travel emissions, which, while small, still represents an environmental impact that we need to account for. We at Grove are committed to leading with transparency and we urge our peers to follow our measured approach to adoption, which we’ve made accessible to all organizations through our partnership with Gravity.”

To quantify and address this impact, Grove has expanded its existing partnership with Gravity, a leading carbon and energy management platform that has developed calculation methodologies around AI emissions. Given primary data disclosures from major AI model providers remain limited, Gravity developed a detailed accounting method leveraging as granular data as possible to assist Grove in assessing their AI-related carbon footprint2. These findings will be published in full detail in Grove’s 2025-2026 annual sustainability report to ensure full transparency, expected to be issued in May 2026. As data transparency increases, Grove and Gravity will continue to evolve their approach to precisely calculate AI-related emissions and share updates so other businesses can do the same.

Grove and Gravity estimate that Grove’s projected 2025 AI-related carbon footprint is 17.8 metric tons of CO2e, which is equivalent to roughly 6% of the Company’s 2024 Business Travel emissions (299 metric tons of CO2e). While the footprint of this AI usage is small relative to Grove’s overall carbon impact, primarily limited to inference and query tasks, the company recognizes the broader environmental risks of AI adoption, including increased energy and water usage, electronic waste and greenhouse gas emissions that contribute to climate change.

“Measuring the impact of the products and services we use has always been an input into understanding the levers a company has to improve its sustainability and operational efficiency,” said Saleh ElHattab, CEO and Founder of Gravity. “With AI’s proliferation and energy intensity, we’ve partnered with Grove to deliver an early blueprint for tracking an organization’s consumption of these new tools.”

Future Goals

Grove says it’s committing to the following sustainability goals:

Measure and integrate AI emissions into overall Scope 3 carbon reporting, in partnership with Gravity;

Invest in nature-based, community led carbon offsets to balance emerging AI-related emissions
Identify opportunities for AI-related emissions reduction;

Opt for large language models and AI tools that prioritize sustainability and transparency whenever possible; and

Call on the larger industry to adopt AI tools responsibly.

Sharing an Open-Source Path Forward for Industry

To encourage industry-wide accountability, Grove and Gravity are sharing this open-source AI emissions formula using a time-based model based on the available science, research and data. This methodology—covering energy intensity, data usage, hardware efficiency and more—will be publicly available as a resource for other organizations to track and reduce the environmental impact of AI adoption, with the formula covering both LLM queries as well as non-LLM AI tools.

Grove’s sustainability initiatives and annual report can be accessed here.