Despite its lofy goals, it is unlikely that the Carney government’s AI strategy will do anything to advance Canadian sovereignty, empower Canadians, or safeguard democracy. Photo courtesy BalticServers/Wikimedia Commons.

Launched by Prime Minister Mark Carney in June 2026, the Canadian government’s National Artificial Intelligence Strategy: AI for All positions AI as central to the country’s economic future and national security. The six “pillars of the strategy” include lofty goals such as protecting Canadians and safeguarding democracy, empowering Canadians, and achieving shared prosperity. But in this roughly 50-page document, the word “environment”—when used in the sense of our natural surroundings that enable our wellbeing—occurs precisely twice (it is used in terms like “fundraising environment” and “information environment” as well). The absence of reference to our environment is significant because just months earlier a major scientific review had shown that “seven of nine planetary boundaries have now been exceeded,” which means “that several of Earth’s life-supporting systems risk crossing critical thresholds, with severe consequences for both ecosystems and societies.”

AI for All, like many other discussions of artificial intelligence, portrays AI as an immaterial technology. But AI depends upon extensive physical infrastructures and global extractive supply chains. The rapid expansion of hyperscale data centres, advanced semiconductor manufacturing, and increasingly powerful computing systems all require enormous quantities of land, energy, water, and minerals. While significant public opposition has drawn attention to the energy and water requirements of data centres themselves, far less attention has been paid to the ecological and social consequences of the mining that makes this infrastructure possible. These harms remain largely invisible due to the nature of extractive supply chains, which externalize their costs across distant, often economically disadvantaged and isolated communities, and make it difficult to trace the end use of materials.

Alongside AI for All, the Carney government plans to bolster its extractivist economy by ramping up the production of critical minerals, which are required to, among other things, build data centres. Canada’s first Critical Minerals Strategy, released in 2022, committed nearly $4 billion to establishing Canada as a global supplier of the minerals required for the green and digital transitions. Since then, the strategy has evolved, with the federal government placing increasing emphasis on economic security, national security, and the role of dual-use minerals in commercial and defence supply chains. Against a backdrop of geopolitical tensions with the United States and intensifying competition over strategic resources, the Carney government now presents expanded mineral extraction as essential to Canadian sovereignty and defence.

The Department of National Defence and Natural Resources Canada are co-leading the government’s participation in NATO’s Critical Minerals High Visibility Project and provincial governments are seeking a share of the spoils. For example, the Government of Newfoundland and Labrador highlights how the province’s “mineral potential aligns with NATO’s defence-critical raw materials priorities, with potential for 9 of the 12 minerals identified as defence-critical.”

AI for All and the Critical Minerals Strategy are not independent policy agendas but mutually reinforcing industrial projects: the expansion of hyperscale AI infrastructure generates demand for critical minerals, while the urgency of securing AI competitiveness provides political justification for accelerating extractive development at the expense of environmental protections and Indigenous rights. The two strategies are further entangled by mining companies using AI in various aspects of their enterprises, from exploring for new deposits to strategic mine planning. Both AI for All and the Critical Minerals Strategy make the ecological and social costs of obtaining these minerals invisible, in part by not talking about these costs and in part by framing extractivism as technologically inevitable and geopolitically necessary.

The material foundations of AI

The servers at data centres carrying out computations and storing information constitute the material infrastructure most apparent to people when they think of AI. But AI, and the closely related “cloud,” requires far more resources than often assumed. Systems like ChatGPT, Claude, and Grok rely on electricity transmission networks, cooling technologies, fibre optic cables, and new electricity generators. As governments and technology firms race to expand AI capacity, the rapid construction of digital infrastructure is generating........

© Canadian Dimension