AI for all can't mean AI for some: Why Canada's strategy must reach beyond the cities

In a small Nunavut community, a digital skills centre was just getting started. It was part of the Katinnganiq Makerspace Network, a Smart Cities winning project designed to open Makerspaces across 13 Nunavut communities. Someone had taken on the role of local champion, the person who showed up, troubleshot the WiFi, and convinced parents and employers that AI training was worth their time. Then they got another job offer and left. The program stopped. Not because the community didn't want it. Not because the technology failed. It stopped because when she left, there hadn't been enough time to train someone to take her place. The community was back at step one.

That story is not an exception in rural, remote and Indigenous Canada. It is the rule. It is the story that Canada's new national AI strategy, AI for All, risks repeating at scale.

Prime Minister Carney launched AI for All on June 4 in Toronto. The strategy targets 250,000 new jobs, AI adoption rising from 12 per cent to 60 per cent of Canadian businesses by 2034, and $200 billion in economic growth. These ambitions assume a Canada with infrastructure and connectivity already in place to absorb them. According to the CRTC, roughly 13 per cent in rural Canadian still lack high-speed internet access - a significant gap compared to urban centres. On reserve, that number falls to 61 per cent, a gap of almost 40 per cent. Rural users who do have connectivity download at half the speed of their urban counterparts.  When access to AI moves at half the speed, or may not even be available, rural Canada's ability to compete is hamstrung before it even leaves the starting block. 

One of strategy's key programs is the refunding of the CanCode program, which has, since 2017, funded coding training for K-12 students and teachers. It is currently the best opportunity to level the playing field for rural and Indigenous students across Canada. Even still, while it funded roughly $9 million under its previous envelope, 80% of the delivery agents funded were based in urban centres. There is an opportunity to improve on that this time around. Creating strong mentors who speak the language of rural communities and have a long-term vested interest in those communities will require stronger funding for the rural organisations building long-term futures and opportunities on the ground in rural Canada.

Rural and Indigenous communities prepare for these contingencies. We know a program's success often depends on the individual who advocates for and administers it, and that creating conditions to support long-term growth is vital to any successful strategy. This is why many organizations, including our own, run "train-the-trainer" models as part of how we work, not just a one-off. Training the next generation of local leaders is the breadth of good community programming, and it rarely fits within a single government program cycle. 

For over a decade, Ampere has supported learners not through one-off programs but across every stage of life, from early exposure to digital skills through to employment, entrepreneurship, and eventually mentorship. When the mentors you look up to speak your language, went to your school, and live the same life you see around you, rural Canadians see themselves as having an opportunity in that field. Look no further than Jordan Tootoo, whose success changed what young people in Nunavut believed was possible in hockey, or the Strumbellas, who did the same for musicians in small-town Ontario. That is the mechanism by which ambition travels. Twenty-five per cent of Ampere's workforce identifies as Indigenous. That is not a credential. It is a consequence of building the right way.

It will be tempting for the funding to flow to easy wins. I often tell funders that while an urban-based digital skills agency will get you millions of light-touch interactions that blend well into an individual's overstocked tech ecosystem, rural-based community organizations can deliver powerful, sustainable, long-term impacts that create lasting legacies for all Canadians. Calling this strategy AI for All is a commitment. Here is what holding it to account looks like: dedicate a meaningful share of CanCode and National AI Literacy funding to rural- and Indigenous-led organizations with long-term presence in the communities they serve. Fund train-the-trainer models that outlast any single program cycle. Measure success not just by the number of Canadians reached, but by whether the communities underserved by every previous wave of digital investment are finally being met where they are.

The infrastructure for a genuinely national AI future is already growing in rural Canada. AI for All needs to mean what it says. Canada can get this right. But only if the strategy travels all the way down the rural route.