Summits |  Meetings |  Publications |  Research |  Search |  Home |  About the G7 Research Group

G7 Environment Meetings | G7 Energy Meetings
Road to the 2025 Kananaskis Summit

G7 Energy and AI Work Plan

Toronto, Ontario, Canada, October 31, 2025

In line with the AI for Prosperity Statement endorsed by the Leaders of the G7 in Kananaskis, Alberta, Canada, on June 17, 2025, we, the G7 Ministers responsible for Energy, concur on a G7 Energy and Artificial Intelligence Work Plan – with the support of international and industry partners at the G7 Energy and Environment Ministers’ Meeting held in Toronto, Ontario, Canada, on October 30th, 2025.

In the face of heightened global challenges and technological competition, we recognize that cooperation among G7 partners, trusted partners, industry, and international partners, to address the energy challenges of AI and harness the innovative potential of AI, is key to the success of our economies' digital transition, as well as broader energy, economic and national security imperatives. We have a shared ambition that AI data centers be reliably and cost-effectively powered through the adoption of technical and market-based solutions, such as optimizing the resource management and demand-side response related to the energy requirements of AI and data centres, in a manner consistent with domestic priorities and applicable international commitments.

We recognize that meeting the energy demands from AI and data centers requires additional energy capacity depending on national circumstances, the optimization of the management of electricity demand and supply, investments in and access to the grid and transmission technologies, energy storage capacity and demand side response/management, as well as critical equipment and materials. We note the critical role of baseload and backup energy, including storage, to meet the energy demand and reliability requirements of data centers. We note that the application of resource-efficient models and tailoring models to tasks are solutions that can reduce costs when used appropriately.

We affirm that investments in energy systems and solutions to power AI must be accomplished in a manner that maintains energy reliability, access, and affordability for our populations. 

In this context, we commit to work with industry, domestic stakeholders, and international partners to:

1. Ensure access to secure, reliable, durable, and affordable energy supply to power AI data centres consistent with domestic priorities and international commitments, by leveraging a wide range of technologies and sources, as well as by advancing cost-effective approaches, such as by strategically locating data centers based on energy ecosystems, as to optimize and stabilize the use of power grids and their expansion.

2. Identify proven and cost-effective solutions to address the energy challenges faced by AI and data centres in collaboration with the IEA Energy and AI Observatory where appropriate, through:

• Varied and scalable approaches and best-practices to build the energy mix and ecosystem required to power data centers, in line with domestic and international commitments,
• Sharing of best practices and of technical and market-based solutions to optimize the management of AI and data centres’ energy demand, including identifying innovative energy solutions that could support the successful delivery and build-up of short- to long-term energy for AI solutions.

3. Promote and enable AI use for energy innovation and breakthrough energy and material discoveries by

• Delivering, where needed, programs and other initiatives to support domestic innovators and the development of AI solutions to accelerate energy and resource breakthroughs, scientific discoveries and innovation – in line with domestic priorities and applicable international commitments.
• Exploring opportunities for knowledge exchange between G7 research institutions, laboratories and centres of expertise.

4. Unlock energy industry competitiveness and accelerate AI adoption through improved trust in AI from consumers, relevant regulators, policy- and decision-makers by:

• Facilitating the sharing of non-commercially sensitive data, information, and use cases to address AI adoption hesitancy in our energy, utility, and resource sectors, as well as promote the responsible deployment of digital transformation and of AI solutions to real-time energy operations.
• Supporting the sharing of proofs of concept of AI solutions and AI-powered processes in energy industry operations while ensuring their safe and responsible application to utilities and end-use sectors.
• Upholding a worker-first approach that does not replace workforces but rather prioritizes AI literacy and skills development to enable them to increase their productivity and realize other benefits related to AI that will make our communities more prosperous and competitive.

5. Promoting best practices for improved understanding and transparency in the energy and resource requirements of AI models and data centre operations, such as through:

• Identifying voluntary industry-led, academic, or multi-stakeholder initiatives, where appropriate, and proven solutions for AI developers, manufacturers and adopters aimed at enhancing the measurement, monitoring and reporting on the energy and resource requirements of AI models and hardware, including their externalities.
• Developing voluntary initiatives and sharing research on how to optimize the costs and resource use of AI models, and data centres, such as the use of compressed models and tailoring of models to their specific tasks.
• Exploring platforms to promote the sharing and cataloguing of open-source solutions, such as open-source models, structured, and trusted non-commercially sensitive data, to expand the adoption and use of AI across G7 energy industries.

6. Build the enabling environment for an effective and secure application of AI solutions across energy systems and operations by:

• Developing a future-ready and AI-enabled workforce by working with international partners to identify effective approaches to up-skilling and re-skilling the existing workforce and attracting and integrating AI expertise into the energy sector workforce.
• Working with relevant international partners and trusted allies to advance the mapping of some infrastructure and supply chain risks for critical equipment and materials needed to build the technology and electricity systems required to power AI and data centres.
• Exploring best practices and proven cases of AI technology uses to streamline or significantly reduce the burden of the regulatory review of energy and resource projects in a trusted manner.
• Demonstrating how AI can lower electricity cost through enhanced integration of energy and optimised grid management, including demand and network flexibility, and improved forecasting of energy supply and demand.

7. Leverage existing platforms, including industry, academic, and multi-stakeholders led initiatives, to promote business-to-business and government and industry information and data sharing on demand-side solutions aimed at optimizing energy use of AI models, data centres, and the grids that power them by:

• Showcasing technological breakthroughs, such as in material and load management, that optimize and improve grid stabilization and optimize the energy and resource use of AI and data centre operations.
• Promoting the sharing of non-commercially sensitive information on the use of technical and market-based solutions to modulate the power demand from data centers.
• Sharing scalable use-cases of circularity in AI and data centre operations to mitigate negative externalities and generate benefits for people and preserve our natural resources, in line with domestic priorities and international commitments.
• Demonstrating the successful use of AI technology in identifying, forecasting, increasing traceability, and mitigating exogenous and endogenous risks to energy and resource operations and supply chains including from malign actors.

8. We note that AI data centers, their components, and supporting energy systems and infrastructure should be secured in a manner aligned with reducing the risk of infiltration or influence. We also encourage the secure and responsible deployment of AI in critical energy infrastructure and energy systems to reduce the risks of, and address, cyber-based threats.

We welcome the endorsement of the G7 Energy and AI Work Plan by Australia, the Republic of Korea, and Ukraine.

[back to top]

Source: Natural Resources Canada

—	This Information System is provided by the University of Toronto Libraries and the G7 Research Group at the University of Toronto.
	Please send comments to: g7@utoronto.ca This page was last updated October 31, 2025.

All contents copyright © 2026. University of Toronto unless otherwise stated. All rights reserved.