According to a new research from Canva, AI-assisted coding is gaining momentum, but quality control and review issues mean that they are still wary of passing the reins to the machine.
The graphic design platform, which votes for 300 decision makers in North America, the UK, Europe and Asia, found that AI adoption is almost universal, with 92% of companies using AI coding tools, and over three-quarters (78%) relying on them every day.
Canva's data shows that this adoption is set to accelerate, with 66% of organizations hoping for a sharp increase in AI coding usage over the next 12 months and hope for even greater productivity (64%).
However, AI has emerged as a powerful catalyst for enterprise software engineering, but decision makers are trying to be thoughtful about their use by quickly prototyping, exploring new ideas, and generating first drafts of valuable code.
It flags almost all (95%) potential risks when deploying AI-generated code without adequate reviews, pointing to the need for guardrails in terms of serviceability, security and quality.
Of those guardrails, the human loop system is probably the most important. To maintain the quality of the code, 93% of companies say that AI-generated code is reviewed all the time or frequently before merging.
Canva said this peer review culture highlights that while AI can generate code in seconds, human judgment is a key gatekeeper of production, and that governance structures enhance surveillance, more than half (58%) of organizations responsible at the CTO and CIO levels show.
These tech leaders have focused on how AI influences developers' skill profiles, and almost all (94%) of AI tools change the nature of their work.
I recommend reading
However, enthusiasm is carefully mitigated by 30% of people who are overreliant on AI without having to take the responsibility of developers, and 1 in 5 (21%) junior engineers worrying about developments occurring.
Canva said these concerns underscore the importance of training to ensure developers not only adopt AI tools, but also to be less visible and accountable.
“Engineering AI is a real, disruptive change that cannot be ignored,” said Brendan Humphreys, CTO at Canva.
“Combined with human judgment and expertise unlocks great benefits, from rapid prototyping to faster development cycles and greater productivity.
“Engineers who thrive in this new era will not adopt AI to enhance and replace thinking and production. They bring creativity, critical thinking and a deep sense of ownership. That's how organizations quickly, innovate and responsibly build.”
