Early on the first day of the CIO Association of Canada’s (CIOCAN) annual Peer Forum conference, Lenovo’s John Encizo asked the room a simple question. How many of you have given your AI agents employee IDs?
One hand went up.
The gap between where the AI conversation has moved and where most organizations actually sit offered a view into Canadian tech that showed how much work still needs to be done. The room full of technology leaders from many of Canada’s largest organizations also showed they’re trying to figure it out together.
And in between sessions, Digital Journal interviewed CIOs across industries who all said pilots are underway, but speed is not always a strategic advantage. More time is needed to consider governance, risk management, business opportunity and creating technology that actually serves humans.
The conference, CIOCAN’s 20th and its return to Vancouver, where the first chapter started, covered topics ranging from agentic AI, data governance, security, sovereignty, and the shifting CIO role.
Three threads ran through the day: the operational foundations that most organizations still haven’t built, the trust deficit between IT leaders and their business counterparts, and a harder question about whether Canada is positioned to capture value from AI or just supply the raw materials.
20 years of duct tape
Encizo told Digital Journal ahead of the conference that organizations have spent two decades layering digital tools over broken processes. On stage in Vancouver, he made the case that agentic AI will scale that breakage at machine speed.
“If you are not asking ‘why’ with every AI project that comes across your desk, you are basically lighting money on fire,” said Encizo, field CTO and director of enterprise AI for North America at Lenovo.
He shared Canadian data from Lenovo’s recently released CIO Playbook that shows organizations are seeing $2.87 in return for every dollar invested in AI which is actually ahead of the U.S. average. But only 18% of organizations have a formal AI policy, and 67% of employees are already using unsanctioned AI tools, referred to as shadow AI
His advice to attendees is to stop trying to police shadow AI and start reading it as a signal. If employees are going around IT to get work done, that tells you something about the tools and processes they’ve been given.
“Don’t be the ministry of no,” he said. “They’re going to use it anyway. Get ahead of it and give them sandboxes.”
Variations on that problem came up throughout the day. Organizations are trying to run AI on top of infrastructure and processes that weren’t built to support it.
James McGregor, chief technology officer at the City of Kelowna, described managing a 30-year technology footprint that still includes applications built on Cold Fusion, a web programming language from the late 1990s that predates most of the modern tools developers use today, and some of the developers.
Alongside that legacy code, his team is building cutting-edge AI chatbots. The city started its AI work in 2018 and processed 180,000 questions through chatbots last year.
That’s not to say there haven’t been hiccups.
When they asked one chatbot which Kelowna beaches have lifeguards, it confidently named several. None of Kelowna’s beaches has lifeguards. The answer wasn’t a complete hallucination. It came from an ancient PDF buried on the city’s website from a time when there must have been lifeguards.
McGregor called his approach “playing small ball.” Get singles and doubles. Avoid swinging for home runs when the failure rate on AI projects runs roughly double that of other technology projects.
CIOCAN president and national chair Shaun Guthrie reinforced the picture, sharing that he arrived at RJC Engineers a year ago to find no data governance, and no one leading reporting. He rebranded his own title from VP of technical services to VP of business technology just to shift how the organization perceived his team.
“You can’t build on top of shit,” said Guthrie, “You have to build a solid foundation.”
Nicole Donatti, energy transformation lead at Data Elephant, described arriving at a client that claimed its data foundation was ready and that it had launched 80 proofs of concept.
Her team mapped them and found they represented about six unique value cases with overlapping data needs.
Encizo had a phrase for that kind of drift earlier in the day: “pilot theatre.”
Why the boss thinks IT is firefighting
Geoff Nielson of Info-Tech Research Group added numbers to something many in the room already felt.
In a session drawing on research Info-Tech has been conducting across thousands of organizations, he presented data from more than 25,000 IT leaders and over 85,000 diagnostics.
The data showed three quarters of CEOs believe their IT function is currently in firefighting mode. For CIOs trying to earn the mandate to lead on AI, that perception is the first obstacle.
Nielson opened by referencing a recent article in The Economist by Ethan Mollick, ‘The IT department: Where AI goes to die.’ The piece argues that senior leaders should keep AI away from IT because IT will just lock it down with governance.
“That should hurt all of us in a very deep and visceral way,” says Nielson, senior vice president at Info-Tech Research Group.
The two biggest drivers of business satisfaction with IT, according to the data, are relationships and delivery capacity. Technology quality ranks lower.
The CIOs who earn the mandate to lead on AI are the ones who have built trust through consistent execution and face time with the business.
Kristin Wilkes, vice president at Long View Systems in B.C. and a former university CIO, drove the point further in a breakout on the evolving CIO role.
She cited a Harvard Business Review finding that 40% of companies have an official AI subscription, but 90% of those same companies report employees using personal AI at work.
“That shadow economy is actually telling you that there is a demand and potential for productivity inside your organization,” said Wilkes, vice president, BC, Long View Systems.
Who captures the value
Chief evangelist at Zoho Raju Vegesna, whose thinking on platform dependency and sovereignty he shared with Digital Journal ahead of the event, stepped past the operational questions and asked something bigger about Canada’s position in the AI economy.
He made his case through the lens of history. Colombia grows the best coffee beans, but Italians captured the value by building the coffee culture. Mexico grows cacao, but the Swiss built a chocolate industry. It has always been the toolmakers who capture the value while raw-material producers watch it leave.
Canada produces a massive energy surplus, but Vegesna says 89% of it flows south.
Energy is the key input for AI infrastructure. Canada has the energy, the talent, and the infrastructure to build its own AI industry. Vegesna challenged those in the room to play a role in building it, rather than just powering someone else’s.
“History remembers builders, not suppliers,” said Vegesna.
Most of the day’s sessions focused on how CIOs should deploy AI.
Vegesna pushed the audience to think bigger. In a room where most hands stayed down when asked about agent employee IDs, there’s clearly work ahead. There’s also an opening for people willing to do the building.
Day two of the Peer Forum shifts to the human side of the CIO role, with sessions on board readiness, personal resilience, and the 2026 threat landscape.
Final shots
- If three-quarters of CEOs think IT is firefighting, CIOs won’t be invited to lead on AI until that changes.
- Data governance, process mapping, and organizational readiness are the work that has to happen before AI can do anything useful.
- Canadian CIOs have a chance to shape where the value from their AI investments lands, inside their own organizations or with the platform companies building the tools.
Digital Journal is the national media partner for the CIO Association of Canada.
