Canva, the design platform, is rewriting the rules of technical interviews. The company now expects candidates for engineering roles to actively use AI tools such as GitHub Copilot, Cursor, and Claude during their assessments, aligning its hiring process with modern software development practices.
The update applies to backend, frontend, and machine learning roles. According to Canva, nearly half of its engineers already use AI tools daily to navigate large codebases, prototype quickly, and write production-ready code. Until recently, however, interviewees were barred from using these same tools.
“Dismissal of AI tools during the interview process meant we weren’t truly evaluating how candidates would perform in their actual role,” the company wrote in the blog post. Furthermore, they noted that the traditional algorithm-heavy assessments failed to reflect the real demands of engineering work, which now includes AI-aided problem solving, code review, and iteration.
Rather than trying to detect covert AI use, Canva has introduced a new interview format called “AI-Assisted Coding,” which encourages candidates to solve product-focused challenges with their preferred AI assistants. The questions are deliberately complex and ambiguous—designed to test a candidate’s technical judgment, comprehension, and ability to refine AI-generated code.
Initial results from the pilot were promising. Candidates who succeeded demonstrated strong debugging skills and strategic AI usage. In contrast, those unfamiliar with AI often struggled, not due to a lack of coding ability but due to difficulty steering the tools effectively.
The company says this shift is part of a broader “AI Everywhere” philosophy. Candidates are informed in advance and urged to practise using AI tools before the interview. Canva mentions that technical rigour remains essential but is now assessed in context, with AI as part of the workflow.
