From pilot to practice: How EY Canada scaled M365 Copilot adoption

Since the beginning of the Industrial Revolution, technological innovation has thrown open doors to new possibilities for business, from the onset of machines and more complex tools to computers, the internet, AI and, most recently, generative AI (gen AI).

While today we can’t imagine our world without them, disruptive technologies have traditionally been viewed with skepticism. Taking the plunge to implement and transform - particularly at scale - can be daunting.

So, while studies found that 55% of organizations have adopted AI - a market valued at $255 billion and expected to explode to $1.68 trillion by 2031 - just over half (52%) claim that risk factors remain a critical consideration when evaluating new use cases.¹

This may explain why, as recently as 2022, only 14% of CIOs surveyed said they were aiming to achieve enterprise-wide AI by 2025.² While eager for efficiency gains and improved customer experiences that safely deliver lasting value, the cost of failure can be a hefty deterrent.

The Fall 2025 EY-Parthenon CEO Outlook Survey found that 30% of Canadian businesses surveyed listed technology disruption and AI integration as the single, most significant challenge to achieving their financial targets - outweighing risks like geopolitical and trade tensions, labour costs and availability, regulatory unpredictability and even supply chain fragility.

And yet, the possibilities continue to dazzle. With gen AI and related technologies, up to 30% of hours currently worked could be automated by 2030.³ AI has the ability to not only solve for business needs but the power to change how we operate, empower people, deliver on customer promise and potentially pave the way to a new approach to conducting business.

To remain competitive, we can’t afford to fall behind on AI - either by moving too slowly or by allowing unsanctioned, “shadow IT” to spread faster than IT can govern it. Because as the saying goes, if you wait until you’re ready, it’s almost certainly already too late.

By August, 82% of our professionals had active licences, with 9 out of 10 people using it at least occasionally. Of those, 67% used the tool at least once a week, with 14% identified as power users who navigated Copilot daily, representing real behaviour change in just months. Daily users reported the tool saved them a minimum of 10 minutes a week in at least one application, that drafting content was faster and that the quality of work improved.

Strong security controls, rapid growth in power users and high satisfaction among active users helped reach steady‑state faster than expected. Along the way, we built playbooks, refined our approach and captured practical lessons that now guide how we scale AI across the business:

1. Change beats hype — with tone from the top. Building strong champion networks with representation from every level of the organization - including our most senior leaders - was one of our biggest accelerators. Users identified peer‑led learning and protected time to experiment as the biggest drivers of adoption, with human-led training rated three times more effective than e-learning. Networks created trust, credibility shared practical tips and helped peers overcome adoption barriers.

But champions alone weren’t enough.

Our Executive Committee spoke openly and often about how they were using M365 Copilot and the impact it was having in every all-hands and leadership call. That visible commitment signalled this wasn’t simply a project, but a strategic priority and a significant shift in expectations around how work gets done across the firm.

2. Governance and trust go hand in hand. Enterprise adoption requires confidence — the kind that comes with clarity and trust. We identified guardrails on data privacy and responsible AI on day one, so employees knew the boundaries.

We trusted our people to apply sound judgment, reinforced through specific guidance in our personal commitments and attestations and learning. This approach allowed us to balance innovation with accountability — supporting employees while protecting data, clients and the business. Trust helped us to unlock creativity and speed’ without sacrificing compliance. 

3. Start small, scale fast. Pilots gave us insight into what worked — and what didn’t — before we scaled. We learned early that the adoption process was not linear; it accelerated as we removed friction, busted myths and celebrated wins.

Three activation waves allowed us to learn, adapt and accelerate without overwhelming our teams. Each wave built certainty and created advocates who pulled the next wave forward.

4. Measure what matters. Success isn’t just about licence counts. It’s about the depth of engagement experienced. We tracked active and power users to understand the behaviour changes and responded nimbly to nuances. When formal training was rated effective by only 21%, we doubled down on what worked best — peer‑led learning and self‑directed experimentation.

Usage insights and feedback loops kept us focused on outcomes, identifying users needed enablement and where we should refine our approach — making adoption a living, learning process.

The message was clear: achieving AI value was an attainable goal. Employing a methodical, practical and data-driven approach, Copilot became a “quality booster”, a go-to for brainstorming and a catalyst for new business opportunities and better client service.

Productivity rose by 10%, with high adopters saving up to 3.2 hours in their weekly work. And the ultimate win? We were able to provide faster, smarter and more efficiently to our clients.

In summary

When Microsoft launched M365 Copilot, EY globally made a bold move: deploying 150,000 licences, which placed us among the top three organizations worldwide. EY Canada built on that momentum as Client Zero, proving we could adopt rapidly while maintaining strong governance in alignment with our culture.

With Copilot studio licences now rolling out, this proactive stance has set the stage for a comprehensive and strategic approach to integration and laid the foundation for broader AI strategy and transformation.

If your team is thinking about M365 Copilot or AI enablement, let’s talk. We’d be happy to share our playbooks, pitfalls, and proof and work alongside your team to roadmap an implementation plan that suits your business.