“In Canada, it needs to be tested, it needs to be proven, it needs to be tried and true,” says Danielle Gifford, Director of AI at PwC.
“Whereas in the US, they’re like, well, that’s a really good idea, and you’re a founder with conviction.”
That appetite for risk, she believes, is a key reason why Canadian companies are lagging in the race to adopt AI at scale.
In a conversation at Inventures 2025 in Calgary, Gifford shared how large enterprises are trying to bring AI efforts under control while contending with cultural hesitation, unclear governance, and the fast pace of technological change. Her insights are grounded in years of work across both startups and big firms, and they reflect a broader reality facing Canada’s innovation economy: many companies have the tools, but not the muscle memory to move fast.
As agentic AI becomes more accessible to non-technical staff, Gifford’s message is straightforward: the barrier is no longer technology. It’s structure, ownership, and mindset.
[Watch the interview in full in the video below]
Scattered pilots and unclear ownership stall enterprise AI
Gifford is quick to point out that AI isn’t new. What is, however, is the pressure to operationalize it.
“Companies actually have a lot of different AI initiatives going on,” she says. “What we’re starting to actually see is, where is the ROI and how do we think about adoption from a large enterprise perspective.”
Rather than one coordinated plan, most large companies have scattered pilot projects and isolated teams, often supported by external consultants. What’s missing, she says, is a unifying governance model that brings those efforts together under one strategy. Companies need to consider not just what models they use, but how they structure teams, hire, train, and optimize around AI.
She points to the finance sector as an example of where deployment is more advanced, largely because of structured data and a history of modelling. Other industries tend to start with back-office applications before risking customer-facing tools.
“You’re not necessarily having a customer-facing chatbot that might be having an issue or selling a truck for $1,” she says. “You can actually work and understand the intricacies of the model on the back end.”
Playing it safe is holding Canada back
Gifford recalls working at two different Canadian startups nearly a decade ago. One built augmented reality tools for the furniture industry, the other focused on cloud platforms for small businesses. Both struggled to find traction at home.
“We could not sell in Canada,” she says. “As soon as we went over to the States, we could sell.”
She doesn’t chalk that up to better tech or sharper pitches. Instead, she points to a deep-rooted difference in how risk is viewed. American customers, she says, were willing to invest in something that might not work. “There’s just a different appetite,” she adds.
That cautious approach shows up in enterprise AI adoption as well. Companies hesitate, or they wait for clear returns or industry benchmarks. In the meantime, faster-moving competitors get ahead.
“Sometimes I think as Canadians, we need to be a little bit more hungry,” she adds. “We need to adopt and I think we also need to be okay with the fact that things might not always work out.”
You can’t scale AI without skilled people
While AI adoption often focuses on tools, Gifford sees people as the most critical part of the equation. Reskilling and upskilling will be essential as companies try to integrate AI into everyday workflows.
“Even for someone in my role, change fatigue is real,” she says. “We’re constantly changing the products, the way we work.”
She believes that traditional corporate training models aren’t enough. Instead, companies need to make learning interactive and accessible. One example she gives is hosting “prompting parties” — collaborative sessions to build skills in a fun and hands-on way.
“Once people actually start to use it and adopt it, they actually find better ways of working with it than you do,” she says.
The arrival of low-code agentic AI tools, especially from platforms like Microsoft, is already changing how work gets done. For employees who understand the business and can navigate basic digital tools, these agents unlock new levels of autonomy.
“As long as you know and understand what your processes are inside an organization,” Gifford explains, “you can actually develop these agents around that to execute on some of the tasks and some of the goals.”
For companies that can combine technical capability with organizational clarity, the opportunity is enormous. But without investment in people and process, the tools won’t be enough. Gifford’s view is a reminder that AI might change what is possible, but leadership still determines what gets done.
“Empower them to do the jobs the best way they can, so that they can focus on some of those higher value tasks,” she says, “not some of the manual, data-driven, day-to-day admin.”
Watch the interview:
