Your competitors are doing this – are you?

Stop reading articles about AI. Start doing something about it.

That may seem like a strange opening for an article about AI. But the single most damaging thing happening in boardrooms and executive suites across Canada right now is not ignorance about artificial intelligence. It is the peculiar paralysis of people who understand perfectly well that this technology is reshaping their industry - and have spent the better part of two years attending webinars, commissioning reports, and forming working groups without actually changing anything.

The numbers make this uncomfortable reading. KPMG's 2025 survey of Canadian business leaders found that 93% of executives report using or piloting AI technologies. Two per cent (2%) say they are seeing a measurable return on their investment. Read that again. Ninety-three per cent (93%) in. Two per cent (2%) getting results. That gap - between the adoption of AI and the realization of its value - is the defining business challenge for Canadian employers today. And it is not going to close by itself.

The window for a considered, unhurried approach to AI transformation has closed. Here is what you need to do.

This week: find your two best bets

Do not convene a task force. Do not engage a consultant. Do not wait for a strategy to organically emerge from below. The evidence is unambiguous: the companies seeing real returns from AI share one common trait — senior leadership picks the spots. Crowdsourcing AI initiatives from across the organization, as most Canadian employers have done, generates impressive adoption statistics and almost never produces transformation.

Before you do anything else this week, identify two - not twenty, two - areas of your business where AI can deliver a clear, measurable improvement within ninety days. The criteria are simple. You are looking for processes that are high-frequency and repetitive, where the bottleneck is human processing time rather than human judgement, and where your data is already reasonably clean and accessible. Invoice processing. First-pass contract review. Scheduling. Generating first drafts of routine client communications. Customer query triage. Demand forecasting. Sales report generation.

Ask yourself three questions about each candidate. Where are our decisions slow, expensive or inconsistent — and could AI make them faster, cheaper, or better? Where are skilled people spending time on work that does not actually require their expertise — and could AI give those hours back for the work that does? Where does poor or slow information cause us to make worse decisions - and could AI surface what we are missing?

The cases that answer yes to all three are your best bets. Pick the two with the cleanest data and the most motivated line manager, and move directly to a pilot. Define what success looks like before you start - not "improve efficiency" but "reduce the time from customer enquiry to quote from four hours to forty minutes." Set a ninety-day clock. Report the outcome transparently, including what did not work.

That pilot is your lever. Not primarily for the efficiency gain, though that matters. For the cultural change that comes when people see AI working in their own context, on their own problems. Nothing shifts scepticism faster than a tangible result that belongs to your team.

This month: talk to your people - honestly

There is a number that should concern every Canadian employer, and it is not in the technology pages. According to EY's research, 89% of employees are worried about what AI means for their job security. Not mildly curious. Not cautiously open-minded. Worried. Some 65% are specifically anxious about being replaced. And nearly half of CEOs have already reported that their employees are reluctant or hostile toward AI adoption.

Canada has its own version of this problem, and it runs deeper than the global average. KPMG's global study with the University of Melbourne ranked Canada 42nd out of 47 countries in trust in AI systems. The same study placed Canada 44th in AI literacy and training. These are not rankings that suggest a workforce ready and eager to embrace AI-driven change. They are rankings that suggest a workforce that needs — and has not received — the support and transparency required to make that change productively.

The hostility your employees may be showing toward AI is not a communications problem you can message your way out of. It is a rational response to real events. Klarna replaced hundreds of customer service staff with AI and publicly celebrated it. Amazon's chief executive sent a company-wide memo acknowledging that AI would shrink the workforce. IBM Canada's own research found that 79% of Canadian office workers are already using AI at work - but only 25% are using enterprise-grade tools their employer has sanctioned. The rest are using personal apps on company time, often because their employer has offered no alternative. That is not adoption. That is improvisation, and it carries real data security and governance risk. According to IBM Canada, shadow AI - the unsanctioned use of AI tools at work - is adding an average of $308,000 per data breach in Canada.

This month, you need to have the conversation you have been avoiding. Gather your teams. Acknowledge that AI is changing work - including theirs. Explain what you are piloting and why. Be honest about what you know and what you do not. And ask them something that most employers have not thought to ask: where do you think AI could help you do your job better?

EY found that 77% of employees would be more comfortable with AI if workers from all levels were involved in the adoption process. CFIB research shows that Canadian businesses investing in AI are 5.4 percentage points more likely to also invest in employee training. People support what they help build. That is not a soft people-management principle. It is competitive strategy.

This quarter: build the capability that will outlast the hype

The AI tools available today will look different in twelve months. The fundamental skill of working effectively alongside AI - knowing when to trust it, when to question it, how to give it useful direction, how to apply human judgement to its outputs - will not change. That is what you need to build, and you need to start building it now.

The gap in Canada is stark. KPMG found that 83% of Canadian employees who use generative AI believe they need to improve their skills to use it effectively. Fewer than half say their organization currently provides adequate training. Meanwhile, 82% of Canadian executives believe they are offering sufficient AI training. That is not a training problem. That is a perception gap - and it means employees who need substantive support are being offered something that is not actually helping them.

The Canadian Federation of Independent Business found that SMEs using generative AI gain more than twice the time they invest — an average of 2.05 hours gained for every 0.97 hours spent. That is a meaningful productivity return, and most Canadian businesses are leaving it on the table. The 45% of Canadian businesses now using generative AI (per CFIB data) and the 51% of Canadian employees reporting AI use at work (per Statistics Canada and CDW Canada data) represent the leading edge. The majority have not yet followed.

What effective capability-building looks like in practice: role-specific training, not generic AI literacy sessions. Your finance team needs to understand which analytical tasks AI now handles and what excellent financial analysis looks like in that context. Your legal and compliance team needs to understand the limits of AI-generated document review. Your salespeople need to learn how to use AI-generated insights to have better conversations, not be displaced by them. Generic awareness sessions generate attendance records. Role-specific training generates capability.

This quarter, you also need to do something that almost no employer has done: update your performance review framework. If you are asking people to work differently with AI but still measuring the same inputs — hours logged, volume of output, processes followed — you are sending contradictory signals. An employee who uses AI to accomplish in ninety minutes what previously took a full day is either your top performer or your next departure, depending on how your system reads that result. Reward thoughtful, effective use of AI. Build it into promotion criteria. If you do not, the employees most capable of leading your AI transformation will take those skills to an employer who will.

Starting today: stop spectating

There is one intervention that has an outsized effect on AI adoption within organizations, costs nothing, and requires no technology budget. It is leadership role-modelling, and most Canadian senior leaders are not doing it.

McKinsey's research on AI high performers found they are three times more likely than their peers to have senior leaders who actively demonstrate ownership of and commitment to AI - not in board presentations, but in how they actually work. When a leader uses an AI tool in their own practice, talks about what they tried and what surprised them, shares a result that saved them two hours on a task they used to dread — they give the entire organization permission to do the same. When that same leader discusses AI at every town hall meeting but visibly never uses it themselves, the gap is noticed.

You do not need to become a technical expert. You need to become a visible learner. Open an AI tool today. Use it for something real - a briefing note you need to prepare, a piece of market research, a board deck section you have been putting off. Notice what it does well and where you needed to push back. Tell someone on your team about it tomorrow. That is the first step. It is available to you right now, it costs nothing, and it matters more than most of the things on your agenda this week.

The governance you cannot afford to skip

Moving fast on AI without building governance alongside it is how organizations create liability, not advantage. IBM Canada's shadow AI data makes this concrete: unsanctioned AI use is already costing Canadian businesses an average of $308,000 per data breach. That number will grow.

Within the next thirty days, you need to establish clear policies on four things. First, which AI tools employees may use and for what purposes - and communicate this clearly, because your employees are already using AI whether or not you have a policy. Second, how AI-generated outputs are reviewed before reaching customers or decision-makers. Third, how data privacy is protected when third-party AI tools are involved - a question with particular relevance to Canadian privacy law under PIPEDA and the incoming Bill C-27. Fourth, what happens when an AI output is wrong, biased, or legally problematic, because it will be eventually.

KPMG found that only 29% of Canadian employees say their employer has a comprehensive AI policy in place - up from 18% the previous year, but still leaving the vast majority of employees without a clear framework for how AI should and should not be used at work. You cannot govern what people do not know exists. Communicate your policies as actively as you deploy your tools.

The cost of waiting

The case for urgency is not that AI will make your business obsolete overnight. It probably will not. The case for urgency is compounding.

The Canadian organizations that began their AI transformation eighteen months ago are now on their third or fourth iteration. Their pilots have become workflows. Their anxious employees have become advocates. Their processes have been redesigned, not patched. The gap between them and organizations still deliberating is not growing at a steady rate. It is accelerating.

Statistics Canada documented that business AI adoption in Canada doubled in a single year — from 6.1% to 12.2% between mid-2024 and mid-2025. The pace is increasing. Meanwhile, IDC research shows that for every dollar invested in generative AI, organizations are realising an average return of $3.70, with top performers achieving $10.30. CFIB data shows Canadian SMEs using generative AI gain roughly twice the time they invest - an average of 2.05 hours gained for every 0.97 hours spent. These are not projections. They are reported outcomes from organizations that started earlier than you and kept going.

The most dangerous position for a Canadian employer right now is comfortable deliberation - knowing enough about AI to feel informed, but not yet doing enough to be competitive. That is precisely where most Canadian organizations find themselves. Ninety-three per cent (93%) have started. Two per cent (2%) are getting results. The organizations in that 2% did not get there by waiting for a more convenient moment or a more complete strategy.

There is also a competitiveness dimension specific to Canada that deserves naming directly. Canadian businesses are not only competing with each other. They are competing with American, European, and Asian firms that are moving faster, spending more, and building institutional AI capability at scale. The question of how quickly Canadian employers act on AI is not only a question of business efficiency. It is a question of whether Canadian businesses remain competitive in an increasingly AI-shaped global economy.

Your employees are already using AI. Your competitors are already using AI. The technology is not waiting for a more convenient moment. The only question is whether you are leading this transformation or being led by it.