Trust is one of the most valuable assets an organization has. And as AI moves into everyday work, that trust is being tested in new ways.
That was the starting point for our recent panel, Creating a Culture of Responsible AI, an intimate conversation between our speakers and a highly engaged audience. Over the course of the discussion, one idea kept coming back: governance alone does not create responsible practice. People do.
Policies, frameworks and controls matter. But they only work when they live inside a culture where people understand the values behind them, feel safe asking hard questions and have the judgment to act well when the rules run out.
The real tension is agency
Why is AI spreading so quickly? The panel pointed to a familiar reason. People have too much to do, everyone is looking for a solution, and AI is right there, easy to access and easy to use.
That pressure is powerful enough that the risks we know about haven't slowed adoption down, even while the biggest risks remain unknown. According to our panelists, the greatest tension between what AI can do and what organizations should do comes down to agency: how much decision-making power we hand over to AI.
That question touches everything. It shows up in AI being used for fraud, in where the ethical line sits on using customer data, and in how to balance the very different interests of employees, customers, partners and shareholders. One panelist also noted that adoption is still early in many places, citing an AI adoption rate of roughly 12% in Canadian workplaces. The decisions organizations make now will shape what comes next.
The encouraging counterpoint: as AI expands, human intelligence becomes more important, not less.
Build a sandbox, with an off ramp
AI will change the workplace. The question is how leaders keep pace without losing sight of their people, customers and values. The panel offered three practical answers.
Make room to experiment, and know when to stop. Organizations need a safe space to try AI, test assumptions and learn what works. But a sandbox needs an exit. Before decisions run through AI, there should be a clear point where someone can say, "This is too much for AI to handle." Good governance isn't only about knowing when to move forward. It's about knowing when to pause or change course.
Invest in AI literacy. Training matters, but so does honesty. Leaders should ask their teams directly: how are you really using AI? The answers are often different from what the policy assumes.
Lead by example. When leaders model how AI should and shouldn't be used, they set a standard people can see, and they make accountability possible.
The panel also put AI in perspective. Technology is only one of many forces reshaping organizations right now, alongside geopolitical and economic uncertainty. That makes it even more important to focus on priorities. Organizations can't control everything, but they can control how, when and why they use AI. The goal is to make that guidance simple enough that every employee knows exactly when to use it.
Above all, adoption is a behavior change, not a technology rollout. Deploying an AI tool is the easy part. Helping people change how they work, make decisions and take accountability is the real work. So the better question isn't "Do we have an AI strategy?" It's "Are we creating the conditions for people to use AI responsibly and well?"
Speaking up starts with psychological safety
As AI becomes part of more decisions, it gets easier to accept its outputs without question. The panel was clear that the antidote isn't a new tool. It's culture.
People challenge AI when they feel safe to. That means creating an environment where employees are comfortable speaking up, asking questions and saying "I'm not sure this is right" without fear of looking slow or difficult. It also means giving them the knowledge to recognize when something is off. When asked what organizations need most, one answer came back three times: education, education, education.
This is where an off ramp and a speak-up culture work together. A clear point for human review only matters if people feel empowered to use it.
Transformation means reinvesting in people
AI has elevated some companies, but in the panel's view it hasn't reinvented any business on its own. It has no 10-year vision. What separates true transformation from cost-cutting is where the gains go. The organizations getting it right reinvest the time AI saves, freeing people from busy work so they can focus on creative thinking and moving the business forward.
The ones that struggle tend to fall into one of two traps: sending every problem to AI, or never committing at all.
One attendee summed it up neatly: we may have moved from FOMO, the fear of missing out, to FOMU, the fear of messing up. With Canada still behind on AI adoption, it's worth asking whether concerns about governance and unintended consequences are crowding out the chance to experiment and learn. Responsible adoption is essential. But standing still carries risks of its own.
The rise of human intelligence
Much of the work AI is starting to automate is the same work where people have traditionally built expertise and judgment. Preparing the next generation of leaders to question, challenge and override AI means doubling down on human skills: critical thinking, sound and ethical decision-making, and the ability to pull real insight from data. Organizations also have to decide what they're ethically comfortable with.
Underneath it all sits learning agility. When AI can inform decisions in seconds, our ability to learn, unlearn, question and adapt becomes the real differentiator. None of us can predict exactly what the future of work will look like, so the goal is to future-proof our skills, not our jobs.
The panel also noted a generational shift. Younger leaders tend to adopt new thinking and new technology with less hesitation than legacy institutions, and traditional career paths are already changing. The challenge for today's leaders is to think 10 or 20 years ahead, not just react to the right now.
What can we control?
The panel left every leader with questions to answer: how will AI change the way we hire, train and manage? There's no single playbook, but one question ties it all together: what can we control? We can't control the pace of technological change. We can control how we respond, how we invest in our people and how we help them navigate what's next.
Responsible AI isn't something you install. It's something you build, through clear values, leaders who model the way and people who feel safe to speak up. Trust is still the most valuable asset an organization has, and in the age of AI it's built the same way it always has been: by people.