When Canva began exploring generative AI, the question wasn't simply which tools to buy or how quickly its team could start using them. The deeper question was what role AI should play in the future of work, and how Canva could make sure its people stayed at the center of that future.
That question has guided Canva through what its team considers a three-stage journey: AI everywhere, AI impact, and AI native, with a new emphasis on AI accountability. Samantha Garrett, AI and Automation Platform Team Lead, and Sean Gillespie, AI and Automation Impact Lead, led a session at Glean:GO 2026 to share what each phase taught them: new opportunities, new challenges, and a steadily deepening understanding that successful AI adoption is about much more than usage metrics.
Canva was founded on the belief that design should be accessible to everyone, not limited to professional designers. As the company met the next major shift in technology, its leaders applied the same philosophy to AI: rather than treating it as a way to replace people, Canva set out to use it as a way to empower, giving its team the tools, knowledge, and confidence to do more meaningful work.
That mindset shaped Canva's earliest AI decisions, and it set a clear expectation: if Canva was going to ask its people to adapt, it needed to invest in their futures and not just in the technology itself.
A holistic approach to AI adoption
Canva's internal AI journey began in October 2023, when the company launched Glean inside Canvaworld, its internal directory. At that point, Glean gave the team a familiar way to search for information and interact with an AI assistant.
However, by April 2024, the pace of change had accelerated. Enthusiastic Canvanauts had begun widely experimenting with other AI tools such as ChatGPT, which then brought leadership to a crossroads: do they restrict access, commit to a single platform, or offer the team a broader set of secure options.
Canva chose the third path, but not all at once. The company first made Gemini, ChatGPT, and Claude available by request alongside Glean, moving fast enough that all three tools went from approval to a company-wide announcement in just six days. With Canva's workforce already 50–60% engineers and broadly tech-savvy, demand outpaced what the team could license. Even so, adoption of the request-based model climbed to just above 90% of Canvanauts.
That demand led to a second, larger decision: rather than continuing to gate access other than universal access to Glean, Canva rolled out universal access to Gemini, ChatGPT, and Claude as well. The signal mattered as much as the access itself—AI was a tool for Canvanauts, not a tool meant to replace them. This reinforced the team's belief that people should be trusted with access to technology they were already eager to use, as long as that access was secure and supported.
The result was rapid adoption: by October 2025, Canva reported 93% weekly active users across its generative AI assistants. But the team quickly recognized that high usage alone couldn't answer the question that actually mattered: was the work getting better?
Measuring real impact
Once AI was available to everyone, Canva shifted its focus from access to effectiveness. They started with an "AI Discovery Week," giving the global team the time to step back, learn, experiment, and reflect, before moving forward with new AI initiatives.
The harder challenge was measurement. On the engineering side, Canva could lean on established, quantifiable frameworks like DORA metrics. On the knowledge-worker side, it was far less straightforward. The team built calculators to translate time saved into dollars, applied a standard 4x productivity multiplier to account for the compounding value of AI-assisted work, and worked with finance to model ROI. It was useful, but it also exposed the limits of spreadsheet-based measurement: knowledge work's value shows up in ways that are hard to quantify, like better decisions, faster experimentation, or stronger customer interactions.
That experience produced a practical, recurring lesson for Canva: adoption is easy to observe, but impact is hard to define, since much of it plays out at the level of personal productivity. The right measurement approach also varies by function. Sales, for example, can trace AI-enabled changes directly to pipeline and revenue; other teams need to look at quality, speed, risk reduction, or opportunity cost instead.
Becoming AI native
As AI became part of everyday work, Canva also asked a more fundamental question: what does it actually mean to be an AI-native company? The answer wasn't to bolt AI onto existing processes, but rethinking how people, teams, and technology work together.
The team's most important insight was that AI usage shouldn't be determined by job title. The relevant question is what activity someone is performing. A senior engineer writing a performance review uses the same chat tools as anyone else in the business; meanwhile, someone in sales might be writing and shipping code to build Salesforce integrations—work that looks a lot like engineering. The line between "technical" and "non-technical" work was becoming less meaningful.
That realization pushed Canva's engineering teams and knowledge workers to collaborate far more closely than before. Each side had something to teach the other: engineers brought experience with code, agents, and technical workflows; business teams brought new applications of AI across research, communications, sales, marketing, and operations. Teams across Canva began collaborating on shared platforms and capabilities, agnostic of role profile.
Greater AI accountability
Canva is also pushing on AI accountability initiatives, helping its team use AI safely, thoughtfully, and in ways that preserve human judgment. The team is focused on three areas:
Human quality. AI should support and enhance people’s thinking, rather than replace it. Canva recognises that AI can make it easy to generate more content, but that doesn’t automatically lead to better work. The focus is on keeping human judgement and oversight central to the creative process.
Financial accountability. As AI pricing shifts from seat-based licenses toward usage-based models, Canva is mindful of using tokens efficiently while continuing to give teams the freedom to experiment and work effectively.
Safety and scalability. As teams build dashboards, connectors, skills, plugins, and autonomous agents, governance needs to evolve alongside them. Canva is focused on providing secure, supported ways for people to use these tools, with appropriate guardrails and visibility as automation becomes more capable.
An ongoing process
Canva's AI journey is one about continuous learning, adjusting course, and staying grounded in human judgement as the technology and the expectations around AI keep changing.
Canva’s team remains focused on the work still ahead: helping people use AI with intention, building safer systems, measuring impact more meaningfully, and protecting the human qualities that make great work possible. By leading with the human experience first, they can also get a much better reaction from their teams rather than leading with just the tools. By putting people first, giving them powerful tools, and helping them use those tools responsibly, they’re taking the right steps towards becoming truly AI native.
