CRM solutions provider Salesforce, alongside AI community platform Hugging Face, AI developer Cohere, and Carnegie Mellon University, announced today the launch of the AI Energy Score, a new benchmarking tool aimed at enabling AI developers and users to evaluate, identify, and compare the energy consumption of AI models.

The new tool comes as the rapid growth of AI by companies globally threatens to challenge corporate carbon emissions reduction efforts, with the IEA estimating that AI-driven data center power usage could double by 2026. Several tech giants, including Microsoft, Google and Amazon, have highlighted increased energy needs to meet the AI-driven growth in data centers as a key challenge to their decarbonization goals.

A recent survey of sustainability professionals by Salesforce found that companies are grappling with balancing the positive and negative aspects of their companies’ implementation of artificial intelligence on their climate and sustainability initiatives, with many worried about the environmental impact of the new technology, while also anticipating benefits ranging from AI’s ability to support supply chain management efforts to ESG reporting and regulatory compliance.

According to Salesforce, the new tool aims to address the lack of transparency about the environmental impact of AI models, establishing a clear benchmark for the energy efficiency of various models, with the company comparing it to the ENERGY STAR label for appliances and electronics.

Salesforce added that by enhancing transparency, the score can drive market preference for efficient models and incentivize sustainable AI development. Salesforce also announced that it will be the first model developer to disclose the energy efficiency of its own AI models using the new framework.

Suzanne DiBianca, EVP and Chief Impact Officer at Salesforce, said:

“Reducing AI energy consumption lowers operational costs, optimizes infrastructure, and enhances long-term sustainability and profitability. We are proud to work with industry leaders to build a more transparent AI ecosystem.”

Key features of the new AI Energy Score include a standardized rating framework for measuring and comparing AI model energy efficiency, a public leaderboard featuring scores for common AI tasks covering 166 commonly used AI models, a benchmarking portal for developers to submit their AI models to be evaluated and added to the leaderboard, and a new energy use label rating AI models with 1 to 5 stars to enable developers to identify and choose more sustainable models.

Dr. Sasha Luccioni, AI & Climate Lead at Hugging Face, said:

“The AI Energy Score represents a major milestone for sustainable AI. By creating a transparent rating system, we address a key blocker for reducing the environmental impact of AI. We’re thrilled to launch this project and look forward to seeing wider adoption.”