The benefits of artificial intelligence are front and centre. The risks barely make the brochure

What do you think about when you hear the term artificial intelligence?

For many Canadians, the first thing that comes to mind is the generative AI chatbots, such as ChatGPT, Claude, Copilot and Gemini, being foisted on consumers and businesses by an aggressive tech industry.

For others, it is the rapid build-out of data centres to power the AI boom. These facilities use as much electricity as entire towns, and they are being constructed at extraordinary speed with little environmental or community oversight.

For some, it is the theft of copyrighted works and private data used to train AI models. Generative AI tools, in this view, are nothing more than plagiarism machines.

And for yet others, “artificial intelligence” evokes risks to our very humanity, threatening to take our jobs, undermine our cognition and mental health, and create existential security risks.

If any of these reactions describe you, you are not alone. Poll after poll finds that the Canadian public is worried about the development and adoption of artificial intelligence, and for a wide variety of reasons.

And yet, many of these concerns are completely ignored in Canada’s new AI strategy.

When Prime Minister Mark Carney announced the plan, he stood on a stage surrounded by doctors. The document itself describes the opportunities to save money in the public health-care system and to deliver better care to Canadians. Other priority sectors identified in the strategy include energy, transportation, agriculture and manufacturing.

In each of these sectors, the strategy highlights the “practical, sector-specific applications” that promise to solve “actual problems” for businesses and governments. Applications like mapping fertilizer usage, diagnosing genetic illnesses and improving energy efficiency. Nary a generative AI chatbot in sight.

Hold on, you may be thinking. That doesn’t sound so bad. But I thought we were talking about chatbots and data centres and job losses?

And here we arrive at the cynical bait-and-switch at the heart of the new AI strategy.

The federal government presents an inoffensive and alluring vision of Canada’s AI-powered future. In this vision, AI is merely a tool that businesses and institutions adopt to improve productivity. The benefits of adoption are assumed to trickle down to the rest of us, including 250,000 new jobs within five years. That’s the bait.

To realize those benefits, the strategy continues, we need a combination of public subsidies to the domestic AI industry, a national AI training program and a huge investment in data centres. Oh, and we can’t enact regulations that risk “smothering innovation.” That’s the switch.

This bait-and-switch is possible because the strategy fails—perhaps deliberately—to define AI. In practice, “artificial intelligence” is a marketing term that encompasses a broad range of often unrelated technologies. As soon as you get more specific about the technologies, the government’s vision starts to make a lot less sense.

It’s one thing to say that “for Canadians to benefit from AI, they must first learn to use it,” as the strategy claims. It’s another to say that “for Canadians to benefit from crop forecasting software in the agriculture industry, everyone in Canada must first learn to use generative AI chatbots.” Despite both falling under the AI umbrella, these are fundamentally different applications.

Artificial intelligence tools may yet offer real value to Canadians, especially the sector-specific applications the federal government is championing.

But Canadians are right to be skeptical of a federal strategy that aspires to indiscriminate AI integration “at school, at work, at home [and] in the community.” We are already experiencing the harms of unregulated AI, and we can see even greater risks on the horizon.

To ignore or downplay those risks and harms in service of the AI industry is disingenuous and irresponsible.

Hadrian Mertins-Kirkwood is a senior researcher and political economist with the Canadian Centre for Policy Alternatives. His research focuses on federal economic, social and environmental policy, particularly climate change, artificial intelligence and industrial strategy. He is a contributor to the CCPA’s Trade and Investment Research Project and Alternative Federal Budget and holds a master’s degree in Political Economy from Carleton University.