How Salesforce’s New System Achieves AI Sustainability
The environmental cost of AI is emerging as a pressing concern for the technology sector, as its rapid adoption drives demand for computing power and energy resources worldwide.
Since data centres are the pillars of AI infrastructure, consuming substantial quantities of electricity and water for processing and cooling and these facilities, the absence of standardised measurements for AI energy consumption has hindered efforts to assess and address these environmental impacts – leading technology companies to have approached the challenge independently, resulting in fragmented reporting and difficulty in comparing the efficiency of different AI models.
Tackling this challenge, Salesforce has launched a new energy rating system called The AI Energy Score.
This system was unveiled at the AI Action Summit on February 10, introducing a standardised framework to evaluate the power consumption of AI models.
Salesforce measures AI energy efficiency through comprehensive metrics
The system, developed with Hugging Face, Cohere and Carnegie Mellon University, arrives as data centre power usage for AI applications is set to double by 2026.
The rating system comprises a five-star scale to measure energy efficiency, accompanied by a public database of more than 200 models across ten AI tasks, including text generation and image creation.
AI developers can submit both open-source and proprietary models through a benchmarking portal and the framework requires developers to measure and report energy consumption, creating a standardised approach to evaluation.
The public nature of these ratings aims to drive market preferences towards more efficient models.
Boris Gamazaychikov, Head of Sustainability AI at Salesforce, says: "We are at a critical moment with the rapid acceleration of both the climate crisis and AI innovation.
“These two trends are interlinked, as we can see exemplified by new gas development slated to support AI data centres."
The initiative has already received recognition from the French Government and the Paris Peace Forum.
"That's why we created the AI Energy Score - to have a standardised, comparable framework to disclose and compare the impacts of AI.
“We hope that this removes a key blocker to a future where sustainable AI is the norm," he adds.
Dr Sasha Luccioni, AI & Climate Lead at Hugging Face, says: "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."
Salesforce's Agentforce platform demonstrates energy efficiency principles
Salesforce has implemented these principles in its Agentforce platform, launched in 2024, which deploys autonomous AI systems across business functions.
The platform enables organisations to build customised AI agents for applications in sales, marketing and data analysis.
Boris explains: "Agentforce is built with sustainability at its core, delivering high performance while minimising environmental impact.
“Unlike DIY AI approaches that require energy-intensive model training for each customer, Agentforce is optimised out-of-the-box, eliminating the need for costly or carbon-heavy training."
The platform has already demonstrated practical applications in waste reduction.
Good360, for example, a charitable organisation, uses Agentforce to allocate donations, preventing six billion pounds of goods from entering landfills annually while reducing 1,000 employee hours.
Salesforce leads industry adoption of energy metrics
Salesforce has become the first AI model developer to publish energy efficiency data for its proprietary models under the new framework, aligning with its broader industry efforts to address AI's environmental impact.
A Salesforce survey indicates that 65% of sustainability professionals believe their organisations must balance AI benefits with environmental considerations – and the AI Energy Score aims to provide organisations with the tools to achieve this.
Such tools include the rating system that has a label generator tool, enabling AI developers to create and share standardised energy efficiency labels for their models and this transparency is designed to facilitate informed decisions about model selection and development.
Suzanne DiBianca, EVP and Chief Impact Officer at Salesforce, says: "Reducing AI energy consumption lowers operational costs, optimises infrastructure and enhances long-term sustainability and profitability.
“We are proud to work with industry leaders to build a more transparent AI ecosystem."
Explore the latest edition of Technology Magazine and be part of the conversation at our global conference series, Tech & AI LIVE.
Discover all our upcoming events and secure your tickets today.
Technology Magazine is a BizClik brand
