AI as a people strategy: what we’ve learned from rolling out generative AI at scale

Generative AI is usually discussed in the language of productivity, risk, or job disruption. But what if, handled responsibly, it could also reduce stress, build confidence, and make work feel more manageable? 

That question was front of mind when I was asked to lead CGI’s AI transformation programme. We wanted to move at pace, but not at the expense of people. Our starting principle was simple: equip colleagues first and let adoption follow. 

Voluntary by design, supported in practice 

Over the last 18 months, as part of a wider transformation effort, we’ve made generative AI available to all our teams, including deploying the enterprise version of ChatGPT. This week we passed 18,000 active users of that one tool alone. Importantly, adopting AI has been entirely voluntary for our people. 

READ MORE: Navigating the deep tech revolution: Why converging breakthroughs are demanding board-level attention

There have been no targets, no “AI mandates”, and no sense that people need to use the tools to prove they’re keeping up. Instead, we focused on positive encouragement and practical support: training, coaching, peer communities and clear (but enabling) guardrails. The aim was to help people experiment safely and confidently, without turning AI into another compliance burden. 

Looking beyond efficiency 

In very short order, we could see material productivity gains. We could measure big return on investment both subjectively, through user surveys, and objectively, in our business metrics. But productivity is only half the story. A transformation that makes work faster, but more stressful, is a poor trade. 

So we ran research focused on the human impact of generative AI. We wanted to understand what was really happening day to day: did AI simply squeeze more output from the same people, or did it genuinely improve working lives? 

The results were reassuring and, in some areas, striking. 

Nine in ten employees reported a positive impact on their health and wellbeing. Many described feeling less stressed and more confident, with noticeable improvements in work/life balance. 

We also saw a strong link to motivation and satisfaction: 

A number of neurodiverse colleagues shared that AI tools reduced anxiety and helped them work more effectively, for example by supporting structured thinking, reducing the friction of writing, and lowering the cognitive load of repetitive tasks. 

Perhaps most interestingly, the strongest reported benefit was an enhanced ability to manage and complete workload. This is a non-trivial outcome in CGI, a company that prides itself on handling complexity. And the benefits weren’t confined to a narrow set of roles: colleagues across job families reported positive impacts. 

A lesson for tech and business leaders 

For leaders, it’s easy to treat generative AI as purely a potential efficiency play, adopting another tool to shave minutes off tasks and reduce cost. 

READ MORE: AI in Action: From Hype to Real-World Results

Our experience suggests a broader opportunity. Used responsibly, AI can help remove friction, eliminate low-value admin, reduce pressure, and create more space for creativity, problem-solving and deeper client relationships. In other words, it can improve performance by improving the human experience. 

That’s why I increasingly see AI as a people strategy as much as a technology strategy. A healthier, more confident and more diverse workforce is also a more engaged and productive one. In a competitive labour market, wellbeing and retention can be just as strategically important as operational efficiency.