At the recent Fast Company Innovation Festival 2025, Duolingo CEO Luis von Ahn told the audience something unusual these days for a tech chief. He stated that artificial intelligence was enabling his employees to perform their jobs more effectively. Imagine that.
Instead of using AI for the sugar high of increasing EBITDA metrics, von Ahn said what more leaders ought to be considering: AI should be regarded as a complementary tool that empowers people rather than supplants them.
Some executives are trying on a similar line for size. Others are not.
During the Brainstorm Tech conference in 2025, Walmart CEO John Furner said that its use of AI (so far) allows the company to scale without cutting jobs. In fact, Furner pointed out the following: “When we look out two years, three years, five years, where I think we’ll be is we’ll have roughly the same number of people we have today.”
Nvidia’s Jensen Huang—the rather obvious pro-AI proponent in the room—pointed out in an interview with CNN that jobs disappear only “if the world runs out of ideas.” It’s a great line when you’re the one selling the chips powering the AI boom. He also said, “Everybody’s jobs will be affected. Some jobs will be lost. Many jobs will be created, and what I hope is that the productivity gains that we see in all the industries will lift society.”
The three CEOs help articulate an essential question for your consideration: Does AI catalyze job growth or ossify the workforce? And if it’s the latter, will there be an inevitable cascade of job losses?
Put differently, is AI the axe or the broom?
And while Marc Benioff, Salesforce CEO, is on record as saying, “This is about humans and AI working together,” he did shrink the company’s support workforce from roughly 9,000 to 5,000 after rolling out AI agents. As he pointed out on the podcast The Logan Bartlett Show, “I was able to rebalance my headcount on my support. I’ve reduced it from 9,000 heads to about 5,000, because I need less heads.”
Is this how reassurance sounds in the C-suite?
Are we to believe that AI will be dressed up as the tidy colleague rearranging the office rather than the rival waiting in the wings? In essence, we must consider whether—in some examples—it is being deployed either as a tool to forestall hiring or as one lurking in the wings, poised for mass redundancies.
When leaders say, “no layoffs as a result of AI,” we should examine the fine print on those early comments. It may be true in the short term, but it may also be code for “We’re not hiring in the future, either.”
It could also mean no cuts in a specific unit or none this quarter. Jobs could vanish in one quadrant of the company yet materialize in another. Whether workers make the AI leap to pick up those newly devised roles will depend on if leaders develop the necessary upskilling bridges to do so.
Canada offers its own set of examples to learn from.
For instance, Statistics Canada has noted in a research paper that exposure to AI in Canadian firms is task-level, not wholesale. Using a complementarity index, StatsCan found three strata of AI penetration so far in the country.
- 31% of workers are in roles highly exposed and less complementary to AI
- 29% are highly exposed and more complementary
- 40% of employees are found in lower-exposure roles
In short, many jobs are evolving, with roughly half of the highly exposed group likely to benefit if leaders did the basics of AI training and role redesign.
Real-world examples in Canada illustrate this point well.
Air Canada, for example, has been rolling out AI decision tools to help manage flight schedules and operations alongside its staff. The company is not seeking to reduce its headcount base; instead, it is providing leaders with more sophisticated tools to make informed decisions, enabling them to handle maintenance disruptions with less chaos. It’s an example of AI-driven augmentation, not elimination.
The auto-parts giant Magna is embedding AI into robotics and quality checks across its plants. That means fewer tedious inspections for people and more human discernment where it counts, including issues like safety, repairs, and exceptions. It is a case where the machine handles the repetition, and the worker preserves oversight.
Compare this to Shopify and Klue, two Canadian companies that have adopted a different approach to Air Canada and Magna.
Shopify has made it clear to its employees: AI is now the default, and managers need to justify why a role should exist at all before they begin the hiring process. In a leaked memo, one phrase encapsulates the point:
“Employees must explain why AI can’t be used before asking for additional resources, like more staff or time. Stagnation is seen as failure, so everyone is encouraged to keep learning and experimenting with AI.”
While that is not a layoff directive per se, it is an example of a hiring brake that will reshape the organizational design of the firm. The jobs may not vanish overnight, but the ceiling will inevitably remain static or expand at a slower pace.
Vancouver-based Klue cut close to half its staff earlier in 2025 as part of what it called an “AI reboot.” As its CEO, Jason Smith, said to BetaKit, “A layoff of this magnitude acts as a kind of forcing function for us as a company—to find efficiencies, one has to look to AI apps and agents.”
It is an example of a wholesale reset of an organization’s operating model, and it shows how quickly headcount can contract when leadership sees AI as a substitute rather than a reinforcement.
What we must question going forward as leaders is whether AI is the axe or the broom. Is it the blade to headcount reductions or the extra appendage to higher performance?
Or is it both?
In essence, is AI the complementary tool leaders choose to help their organizations and employees become more effective and efficient? Or will AI be used to justify smaller teams and inflated margins?
I hope there will be a future case where more senior leaders use AI to enhance job tasks and make work more manageable for all.
The technology remains indifferent.
The choice sits squarely in the executive chair, and it will be obvious soon enough who treats AI as reinforcement and who deploys it as replacement.
As the curlers might say that are sweeping the broom, “Hurry!”