Karsten Wenzlaff, Advisor
August 26th, 2025
Karsten Wenzlaff is a Crowdfunding and Fintech Expert from Germany. He is Secretary-General of Digit[...]
AI Energy Score Ratings A Step Towards Transparency in AI
AI | Feb 26, 2025
Image: AI Energy Score Label (salesforce.com)
Canada's Cohere Helps Drive AI Energy Score for Sustainable AI
Electricity, data centers, and energy infrastructure oh my. That's right. As the demand for artificial intelligence (AI) energy grows with adoption, sustainable and transparent AI coalitions like the AI Energy Score initiative could become a critical standard in measuring the energy efficiency of AI models, empowering individuals, businesses, researchers and policymakers make informed decisions on AI usage.
By 2028, AI data center electricity consumption could consume up to 12% of the total electricity in the United States, a hockey stick increase from 4.4% in 2023. While the U.S. is planning to add 46 gigawatts of new natural gas capacity by 2030 to help meet this demand (equivalent to the entire electricity requirements of Norway), industry and governments globally are recognizing the urgency for energy capacity and making AI more energy efficient.
Introducing AI Energy Score
A coalition of industry leaders including Salesforce, Hugging Face, Canada's Cohere, and Carnegie Mellon University launched the AI Energy Score initiative, to provide a standardized way to measure and compare the energy efficiency of AI models. AI Energy Score evaluates AI models based on their power consumption when making predictions or generating responses called 'inference' (not when training LLMs) and assigning each a star rating of 1-to-5.
A public leaderboard ranks over 166 AI models across categories such as text generation, image generation, summarization, automatic speech recognition, and object detection, so users can compare energy efficiency across different tasks.
Developers can also submit their models for evaluation through a benchmarking portal.
And as you may expect, similar to ENERGY STAR labels for household appliances, AI models that meet high efficiency standards receive recognizable energy-use labels.
How the Rating System Works
The AI Energy Score evaluates models across ten AI tasks and energy use is measured in watt-hours per 1,000 queries to establish a standardized assessment. Each model is then assigned a star rating based on its efficiency. The top 20% most efficient models receive a five-star rating, while the least efficient 20% receives one star. The leaderboard is updated every six months to reflect new advancements and ensure the ranking system remains relevant.
Cohere’s Involvement and Canada's Action
Toronto based startup Cohere, an AI company specializing in large language models for enterprise is the only known Canadian firm participating in the AI Energy Score initiative at this time, signalling Canada’s growing awareness of the need for AI energy efficiency. Note Canada has built a strong reputation in AI research through institutions like Mila in Montreal, the Vector Institute in Toronto, and Amii in Edmonton.
See: Canada’s AI Competition Report Faces Big Tech Challenges
The Canadian government has started recognizing the challenges posed by AI’s energy consumption. The Standards Council of Canada (SCC) recently emphasized the need to balance AI innovation with sustainability and data integrity. In a 2024 policy brief, SCC warned that AI’s growing energy demand could hinder Canada’s climate commitments if left unchecked. In response, the government has proposed a $15 billion incentive package to attract energy-efficient AI data centers to Canada. It has also aligned with international AI governance efforts, such as the Global Partnership on AI (GPAI), seeking to establish sustainability standards for AI development.
Canada is also moving on legislation. The Artificial Intelligence and Data Act (AIDA, part of Bill C-27), focused on regulating AI transparency and accountability. However, Canada has not yet formally joined initiatives like the AI Energy Score or the Coalition for Sustainable AI.
Implications of AI Energy Rating Standards
If energy ratings for AI models become widely adopted then all stakeholders will start prioritizing efficiency. In practice this might mean companies would begin selecting AI models based on their energy usage (similar to how businesses evaluate their carbon footprint today when deciding certain policies). During the transitional period of adoption to new ratings, companies building AI products while optimizing for energy efficiency could use that as a competitive advantage.
Investors may also begin favouring AI firms that disclose energy consumption data as part of their support for transparency and sustainable AI, unlocking new funding opportunities.
See: Will Nuclear Fuel the Data-Driven Future?
Governments may introduce regulations requiring AI models to meet minimum efficiency benchmarks to ensure AI innovation and development efforts align with broader climate goals.
Will Canada Take a Leadership Role?
Canada must decide whether to help lead in AI energy transparency or risk falling behind. While Cohere’s participation in the AI Energy Score is a positive step, there is currently no national strategy on AI energy ratings. The AI Energy Score represents one of the first major efforts to standardize energy transparency in AI and its adoption would change how businesses, governments, and investors evaluate AI models. It's often said that Canada has the opportunity to lead in sustainable and responsible AI but it will require deeper engagement from policymakers, industry leaders, and research institutions.
The National Crowdfunding & Fintech Association (NCFA Canada) is a financial innovation ecosystem that provides education, market intelligence, industry stewardship, networking and funding opportunities and services to thousands of community members and works closely with industry, government, partners and affiliates to create a vibrant and innovative fintech and funding industry in Canada. Decentralized and distributed, NCFA is engaged with global stakeholders and helps incubate projects and investment in fintech, alternative finance, crowdfunding, peer-to-peer finance, payments, digital assets and tokens, artificial intelligence, blockchain, cryptocurrency, regtech, and insurtech sectors. Join Canada's Fintech & Funding Community today FREE! Or become a contributing member and get perks. For more information, please visit: www.ncfacanada.org
 |  |  |
|---|---|---|
 |  |  |
Leave a Reply