Although AI-assisted coding is gaining momentum, quality-control and review issues mean tech leaders are still wary of handing the reins to machines, according to new research from Canva.

Polling 300 decision makers across North America, the UK, Europe and Asia, the graphic design platform found that AI adoption is nearly universal, with 92% of firms using AI coding tools, and more than three-quarters (78%) of developers relying on them every day.

Canva’s data shows this adoption is set to accelerate, with 66% of organisations expecting their use of AI coding to increase sharply over the next twelve months, fuelled by a desire for further productivity gains (64%).

But while AI is emerging as a powerful catalyst in enterprise software engineering, with rapid prototyping, exploration of new ideas, and generating first drafts of code proving especially valuable, decision makers are trying to be thoughtful about its use.

Almost all (95%) flagged potential risks when deploying AI-generated code without sufficient review, pointing to the need for guardrails around maintainability, security, and quality.

Of those guardrails, human-in-the-loop systems are perhaps the most critical. To maintain code quality, 93% of firms said that their AI-generated code is always or often reviewed before merging. 

Canva said that this peer review culture highlights that while AI can generate code in seconds, human judgment remains an essential gatekeeper to production, with governance structures reinforcing oversight, illustrated by more than half (58%) of organisations placing responsibility at the CTO and CIO level.

These tech leaders are keeping a close eye on how AI is impacting developer skill profiles, with nearly all (94%) believing that AI tools are changing the very nature of the job. 

Recommended reading

• Report: 65% of Devs Say AI Code Needs Major Fixes
• Half of Developers Think AI Can Code Better Than Most Humans
• AI Saves Developers 10+ Hours A Week – But Inefficiencies Cost Them The Same

However, any enthusiasm is being tempered by caution, with 30% citing over-reliance on AI without developer accountability as a top concern, and one in five (21%) worrying that junior engineers may see their development stunted. 

Canva said that these concerns highlight the importance of training to ensure developers not only adopt AI tools but also use them discerningly and with accountability.

“AI in engineering is a real disruptive change that can’t be ignored,” said Brendan Humphreys, Canva’s CTO. 

“When paired with human judgment and expertise, it unlocks significant benefits, from rapid prototyping to faster development cycles and greater productivity. 

“The engineers who will thrive in this new era are adopting AI to enhance their thinking and output, not replace it. They’re bringing creativity, critical thinking, and a deep sense of ownership. That’s how organisations will move fast, innovate, and build responsibly.”