Investor focus is shifting from companies building artificial intelligence infrastructure to those applying the technology directly to improve efficiency, revenues and margins. The transition could play an important role in addressing Canada’s long-standing productivity challenges, particularly in service-based industries.
BNN Bloomberg spoke with Kevin Khang, head of global economic research at Vanguard, about why he sees AI entering a new phase, how it may reshape economic growth and why Canada could be positioned to benefit over the long term.
Key Takeaways
- AI is moving into a new phase focused on implementation rather than infrastructure spending.
- Productivity gains are expected to come from automating tasks and enhancing human labour, not replacing it entirely.
- Canada’s service-heavy economy may be particularly well suited to benefit from AI adoption.
- AI-driven productivity gains could help offset demographic pressures from an aging workforce.
- While risks remain, broader AI adoption could provide a long-term economic tailwind for Canada.
Read the full transcript below:
ANDREW: For more than a year, tech companies building the infrastructure behind artificial intelligence — providers of massive computing capacity — have been the focus for many investors. Attention, however, is starting to shift toward companies that can implement AI in their operations.
Our guest says that shift could bode well for some of Canada’s publicly traded companies and may help improve the country’s lagging productivity.
Let’s get more from Kevin Khang, head of global economic research at Vanguard. Kevin, it’s great to see you. Just to remind viewers, Vanguard is one of the world’s largest ETF providers, among other products.
I was struck by this: you’re not involved in the Canadian business day to day, but you have an ETF — the FTSE Canada All Cap Index ETF — with almost $13 billion in assets and a management expense ratio of about one-fifth of one per cent. You’re clearly known for low-cost products, especially when compared with mutual funds charging more than two per cent.
You’re in the camp that says we’re moving into a second phase of AI, with the focus now on companies that can implement it profitably.
KEVIN: Good morning. And before we move on, I would just say there’s one phrase that summarizes it all, and that’s the “Vanguard effect.” Whenever Vanguard moves into a space — because we care about investor outcomes and long-term success — that tends to create positive externalities for end investors.
Going back to your original question, we’ve been saying for a while that AI is one of those epochal technological advancements that don’t come along every generation. When we first saw its capabilities, our view was that it has era-defining potential — to boost productivity, transform how the economy works and change how companies generate profits.
We think of this as unfolding in phases. Our baseline view is that AI is on track to deliver that transformative productivity boost in the U.S., Canada and many other economies. The phase we’re in right now is reminiscent of earlier technological cycles, such as electricity or mass manufacturing after the Second World War, where there was a deepening of physical investment across the economy.
It’s somewhat ironic. AI is the ultimate cognitive power tool, yet it requires a historic capital investment buildout. That has generated excess demand for what we might think of as old-economy sectors — industrials, materials and energy — and that has shown up in markets like Canada’s. We’re seeing that both in the U.S. and in Canada.
ANDREW: You’re a financial economist by training and have done a lot of research on active portfolio construction and household finance. Have you spent much time looking directly at AI itself? And even if you haven’t, what has surprised you most about the surge of excitement around AI from an economist’s perspective?
KEVIN: This goes back to the idea that AI is a once-in-a-generation development, and its impact will be multifaceted. It won’t all happen at once. There will be multiple innings and phases.
The way AI is diffusing has created something of a binary reaction. Some people focus on labour market implications and worry about dystopian outcomes — whether people will still have jobs. Others are at the opposite end of the spectrum and see a kind of nirvana, where everyone becomes vastly more productive.
Our view is more measured and somewhere in between. AI has significant potential. If you look at almost any job and break it down into tasks, there are many areas where AI can meaningfully boost productivity by automating routine work and helping human workers perform better.
At the same time, there are many judgment-based tasks that remain beyond the scope of current AI. When you combine that with how labour markets have evolved during past technological revolutions, we think this time won’t be entirely different. It will be disruptive, with some sectors affected more than others, but the productivity gains are likely to outweigh the downside.
ANDREW: To close, I know you and your colleagues have studied AI-driven productivity in the U.S. economy. If we switch the lens to Canada, you’ve warned about demographic pressures — particularly an aging workforce.
You’ve also said there’s a risk AI fails to meet expectations. So is there a realistic prospect that AI helps improve Canada’s poor productivity performance?
KEVIN: The short answer is yes. I’d make two points. First, demographics — an aging population and slower labour-force growth — are a long-term reality unless there’s a major shift in immigration policy. Economies facing that reality can’t rely on labour growth for expansion.
Second, Canada still derives most of its income from services. Despite headlines and the TSX’s heavy weighting toward resource sectors, the bulk of economic activity comes from highly labour-intensive services. That makes Canada particularly well positioned to benefit from AI-driven productivity gains.
This is coming at an important time. As demographic pressures intensify, labour shortages become more likely. AI-based productivity gains could help offset that constraint.
It’s not guaranteed — there’s still a lot that needs to go right — but thousands of startups around the world, particularly in the U.S., are working to unlock that potential. From our perspective, AI could become a meaningful and helpful tailwind for Canada’s economy.
---
This BNN Bloomberg summary and transcript of the Jan. 26, 2026 interview with Kevin Khang are published with the assistance of AI. Original research, interview questions and added context was created by BNN Bloomberg journalists. An editor also reviewed this material before it was published to ensure its accuracy and adherence with BNN Bloomberg editorial policies and standards.
