Canva has officially changed its technical interview process by introducing AI-Assisted Coding, and even candidates have expected to use devices such as Gitab Copilot, Cloud and Cursor during interviews for engineering roles. This step is a major change in how the company assesses software engineers, align the hiring process with its in-house developers equipment that already uses on a daily basis.

Till now, Cainwa’s coding interviews, especially for backand, frontnd and machine learning rolls, were much more dependent on traditional computer science fundamental assessment. These tests, which included solving algorithm-based problems without any external help, was no longer a reality of day-to-day in Canva as engineers now regularly depend on the AI ​​tool to navigate large codebase and accelerate development.

Realizing that the disconnect has limited the company’s ability to assess the correct-school performance, Cainwa has now introduced what he says to “A-A-A-Assisted Coding” interviews. Candidates are not only allowed, but also encouraged to use AI during the interview. The company says that the idea is to understand how well they can cooperate with the AI ​​tool, how effectively they break complex problems, refine machine-borne code, and make sound technical decisions.

Talking about the tests conducted for this, Simon Newton, head of canva platforms, said, “Our AI-Assisted interview feels more attractive for both candidates and interviewers, and they are providing a strong future signal about the candidate’s performance. The most important thing is that they are helping us identify the engineers that AI can make AIs thoughtfully and effectively.

This change comes amid growing debate around the use of generative AI in hiring. While some companies look at AI as cheating in the interview, Canva says it is trying to ban the use of AI, the opposite in an industry where such devices are already an integral part of development workflow. Instead, the company is giving priority to transparency and encouraging candidates to use AI openly during the interview.

The company revealed that candidates are now judged on how they guide AI, evaluate the suggestions of AI, and maintain production-level standards in their final solution. Interviews also test how well the candidates debugged the AI ​​code and dealt with vague product requirements.

Newton said, “Interestingly, candidates with minimal AI experiences often struggled, because they could not code, but because they lacked decision to guide AI effectively or identify its suggestions.” “We believe that the future belongs to engineers who can original human creativity and decision with AI capabilities,” he said.