Canada ranks near the bottom globally for AI trust — closing that gap is HR's job

Prime Minister Mark Carney said something at Thursday's AI strategy launch that should stop every HR leader mid-scroll. "Canada ranks near the bottom of countries in AI training, literacy and trust," he acknowledged.

He framed it as a challenge for government to address. He is wrong about who owns it. 

The federal "AI for All" strategy commits $2.3 billion to infrastructure, SME adoption programs, computer access, and youth literacy training. What it cannot fund is the conversation between a people manager and an employee who is convinced their role is disappearing. What it cannot mandate is the governance structure that makes an employee feel safe enough to use AI openly rather than hiding it. What it cannot build is the institutional trust that turns a nervous workforce into a productive one. 

"We need to treat adoption of AI as a change just like we would the adoption of an ERP system or the integration of an acquisition," Scott McAllister, CEO at Prosci, a change management solutions provider, told Canadian HR Reporter in 2024.

"For the longest time, as professionals, we almost focused exclusively on getting the technical solution. And we thought, if the solution is strong enough, people will just naturally adopt and use it. And through a lot of lived experience, we understood that that's not how things happen in the real world." 

Carney's strategy is a technical solution document. The adoption problem is a people problem. That is HR's terrain. 

The numbers behind the nervousness 

The data underpinning Carney's admission is more striking than his brief acknowledgement suggested. According to the Ipsos AI Monitor 2026 — a 32-country survey of 23,532 adults conducted between March and April 2026 — 67 per cent of Canadians said AI makes them nervous, one of the highest rates in the world. Just 37 per cent believe AI has more benefits than drawbacks, near the bottom of all 32 countries surveyed. Only 26 per cent think AI will improve the Canadian job market. Only 20 per cent believe it will improve their own job. 

The productivity figures tell a parallel story. While 62 per cent of workers globally said AI saved them time in the past 12 months, Canada's figure was just 46 per cent. Yet more than half of Canadian employees are now using generative AI at work —  up from 46 per cent in 2024 and just 22 per cent in 2023, according to KPMG's Generative AI Adoption Index. 

That combination — rising covert adoption, low institutional trust, and a productivity gap — is precisely the environment that produces governance failures. Over half of workers globally admit they conceal their use of AI at work, often presenting AI-generated output as their own, according to a 2025 University of Melbourne and KPMG study of more than 48,000 people across 47 countries.

"The responsible use and governance of AI is not keeping pace with its adoption," the report concluded. 

In Canada specifically, 31 per cent of employees are seen as resistant to agentic AI  — nearly double the global rate of 16 per cent. More than half of those resistant employees attribute their hesitation to trust and ethical concerns, with nearly 40 per cent pointing to job security fears.

Lewis Curley, partner in the people and change practice at KPMG Canada, identified the core issue for HR leaders: "When organisations are looking at upskilling their workforce, they must have a clear picture of how their people can use AI agents in ways that deliver meaningful impact on the business. A skilled employee can build an agent to automate tasks and get work done faster, but if it's not being used on work that produces results, companies won't see the returns they expect." 

Why the anxiety is rational

The instinct in many organisations, when confronted with employee AI anxiety, is to run an awareness campaign. This is insufficient — not because communication is unimportant, but because the anxiety is based on accurate information. 

Employees are reacting to AI with knowledge-hiding and job insecurity behaviours that undermine collaboration and innovation, according to research by Ann Schlotzhauer, assistant professor of psychological science at Missouri University of Science and Technology. The rapid pace of technological change and increasing AI use is "intensifying employee job insecurity, making the issue a pressing concern for HR." Workers are not irrational. They have read the same labour market data as their managers. 

Jean-Nicolas Reyt, a management professor who has studied AI disclosure behaviour in Canadian workplaces, put the trust dynamic directly: "A lot of people don't trust their managers to have their best interest in mind, and so to them, why would they share that they're using ChatGPT that tremendously increases their productivity?"

 Employees who are more productive with AI worry that transparency will lead to higher expectations, tighter targets, or arguments for headcount reduction. "If I were to tell you that now I can be 50 per cent more productive, the last thing I would want is for my manager to know about it, because if my manager knows about it, I'm screwed," he says. 

That anxiety intensifies when early-career workers look at the broader labour market signals. "If you think about a lot of the headlines and the storytelling about AI, it touches on junior or entry-level task automation," said Susan Zettergren, chief people officer at Capital One Canada, in Canadian HR Reporter's coverage of AI's impact on workforce culture. "You can quickly see how someone new to the workforce might feel very scared." The response from employers, she argues, must go beyond technical training: organisations need to combine AI skills development with soft skills and collaboration training to preserve their pipeline of future leaders. 

How Canadian HR leaders are getting it right

The most useful evidence for HR practitioners is not from surveys or economists. It is from the organisations already navigating this. 

Carolynn Ryan, senior vice-president of people and CHRO at BC Hydro, described in Canadian HR Reporter's recent roundtable on AI in Canadian workplaces how the utility approached adoption without waiting for employee confidence to arrive on its own. BC Hydro focuses on real use cases — for example, how an HR business partner can use AI to prepare for a difficult employee conversation, or how a people manager can apply it to performance management or goal-setting.

"One of the key things is that air of experimentation and knowing that just like people can make mistakes, AI can make mistakes too," Ryan says. "So, being open with that idea that it's not going to be perfect, but try different things and talk about it. I think confidence comes from that critical thinking and experimentation — not necessarily in trying to master or take a course in it." 

Karen Bateh, vice-president of people and culture at CBC/Radio-Canada, described in the same roundtable how the broadcaster built governance that employees could see rather than just be told about.

"It's obviously within HR but there's a lot of experimentation in our field," Bateh says. Those guidelines were developed in collaboration with various functions, including employee resource groups, unions, and bargaining agents — an approach Bateh says was essential to building legitimacy.

"Let us share openly things that didn't work because, for us, part of our people strategy as part of our broader corporate strategy is… change is this constant that we need to continue to build this resilience and comfort with. And with failure comes learning," she says. 

Both approaches share a structural feature: they start with visible governance and real use cases, not with tools mandated from above. 

Sara Cooper, chief people officer at Jobber, offered a complementary perspective on how HR can get ahead of the shadow adoption problem.

 "We know that managers or people leaders, instead of coming to us for advice, can just tap into ChatGPT," she told Canadian HR Reporter. Rather than attempting to stop that, Jobber's people team is building an internal alternative — a tool that embeds the organisation's own policies and values.

"We are putting in all of our policies — everything around surrounding people at Jobber and how we want to show up as people leaders." The principle: if employees are going to use AI for people management decisions anyway, give them a version that reflects your organisation's values rather than a generic chatbot. 

Holly Ackert, executive director in people and change consulting at KPMG, frames the maturity question clearly for HR leaders trying to assess where their organisation sits. AI adoption, she says, has two layers: there is first "a push to solidify the HR technology foundation," and then, for more mature organisations, the question of "how is AI and strategic AI adoption now being embedded in our processes?"

Most Canadian organisations, she suggests, are still on the first layer — and trying to build trust for the second layer before the first is stable is where the failure mode sits. 

Three things that actually close the trust gap 

Research and practitioner evidence consistently point to the same factors. 

Replace vague reassurances with specific governance that employees can see. Teresa Scassa, Canada research chair in information law and policy at the University of Ottawa, told Canadian HR Reporter that employers need to think on two levels: official tools brought in through procurement, and shadow use by staff. On the procurement side, employers must align contracts and configurations with their legal and security obligations. On the shadow side, structured dialogue with workers — before and during adoption — is one of the most practical safeguards available. Employees who can see a governance structure are less anxious than employees who have been told not to worry. 

Train before you demand, not after you deploy. Lisa Stam, founding partner at Spring Law in Toronto, put this with unusual directness in Canadian HR Reporter's recent piece on AI and the law: "If you throw all kinds of tasks at someone that they haven't been trained for, and then you fire them because they fail to do the job, then you're looking for trouble. It's a new frontier right now, and so anybody who wants to evolve their workplace to stay relevant and have skilled employees, they can't go on the market to find those employees, they don't exist yet. All of us benefit from just training our own existing employees, because it's all new for everybody. There's no AI experts out there." The legal risk is real. So is the practical one: organisations that deploy AI without training create the shadow adoption and error patterns that destroy trust faster than any communications campaign can rebuild it. 

Measure trust, not just adoption. The KPMG data shows that 51 per cent of Canadian employees using AI report improved productivity — but 93 per cent of Canadian enterprises that adopted generative AI report only two per cent seeing the returns they expected. That gap is not a technology problem. It is a governance and trust problem. Most organisations track tool usage. Fewer track whether employees trust the tools they are using, whether they trust that AI will not be used against them in performance reviews, and whether they trust that their manager has a plan for their career in an AI-enabled team. 

Carney's strategy is a genuine and ambitious document. Its job creation targets, compute investment, and SME adoption supports are real signals of intent. But the strategy's own data — Canada near the bottom globally for AI trust and literacy — describes the binding constraint on every target it sets. Moving business adoption from 12 per cent to 60 per cent by 2034 requires a workforce willing to actually use the technology. Building that willingness is not a federal programme. It is a daily management practice. It is a hiring and onboarding decision. It is a collective agreement negotiation. It is a performance conversation. 

It is, in other words, HR.