// round guide · all AI-era interview guides

Canva's AI-assisted coding interview

Canva is one of the few companies that has written publicly about its AI-assisted interview, so there is less guessing than usual. You use the AI tools you normally use, on a problem too big and too vague for one prompt, and you are judged on how you steer the tools and what you do with their output.

Last updated

Practice this round, no account needed

What is known about the real round

Backend, frontend and machine learning candidates are expected to use AI tools in technical interviews.

Canva announced this on its engineering blog in June 2025, naming Copilot, Cursor and Claude as examples. The post says candidates are still assessed on fundamentals and have to own and explain any code they submit, whoever wrote it.

Stated by the company. Source: Canva engineering blog: "Yes, you can use AI in our interviews. In fact, we insist". Last checked October 27, 2026.

Formats change between hiring cycles. Ask your recruiter what your round will be, and see the policy tracker for other companies.

Format overview

In a June 2025 post on its engineering blog, Canva said it now expects backend, machine learning and frontend engineering candidates to use AI tools such as Copilot, Cursor and Claude during technical interviews.

The post says a new AI-assisted coding competency replaces its traditional computer science fundamentals screening for backend and frontend roles, and that fundamentals are still assessed within the new process.

Problems are deliberately more complex and ambiguous, so they cannot be solved with a single prompt. The example Canva gives is building a control system for aircraft takeoffs and landings at a busy airport. Candidates are told about the format in advance.

What is evaluated

  • Knowing when and how to use AI tools effectively (stated by Canva).
  • Breaking down ambiguous requirements (stated by Canva).
  • Making sound technical decisions (stated by Canva).
  • Finding and fixing problems in AI-generated code, and judging whether it meets production standards (stated by Canva).
  • Owning and explaining the code you submit, whoever wrote it (stated by Canva).

Example task types

  • A realistic, open-ended product problem rather than a textbook algorithm.
  • Clarifying requirements before building.
  • Reviewing and debugging what the AI tool produces.

How the practice round runs

Placed runs its own version of this format with a voice interviewer. The code and the tasks are Placed's, not Canva's.

  1. Step 1
    Clarify the problem

    In the Placed version you get a task with gaps in it. Ask the interviewer what the inputs look like and what counts as done before you open the assistant.

  2. Step 2
    Build with the assistant

    Use the assistant for well-defined pieces. The report records what you asked for and whether you checked the answer.

  3. Step 3
    Defend the result

    The interviewer asks why the code works and what would break it. You are scored on explaining code the assistant wrote as if you wrote it.

What gets scored in practice

Every Placed round ends with a report that scores each rubric criterion from 0 to 10, including how you prompted the assistant and checked its output. On Pro, each score comes with the moment in your session it is based on. How grading works.

Where people lose the round

  • Starting to prompt before you understand the problem.

    Canva's post says strong candidates asked clarifying questions first. Spend the first minutes on requirements.

  • Asking the tool for the whole solution.

    Give it well-defined subtasks you can check, and keep the design decisions yourself.

  • Accepting output without reading it.

    Read every block, run it, and say what you checked. Finding the tool's mistakes is part of the score.

A prep plan: three sessions before the real one

  1. Use the AI tool you will bring for a week of real work, so the interview is not the first time you steer it under pressure.
  2. Sit two single-problem AI rounds on Placed and compare the Output Validation scores.
  3. Practice a code review round: saying what is wrong with generated code out loud is the skill Canva names.

Questions people ask

Can you use AI in Canva interviews?

Yes. Canva wrote in June 2025 that it expects backend, machine learning and frontend engineering candidates to use AI tools like Copilot, Cursor and Claude during technical interviews.

Does Canva still test fundamentals?

Canva says computer science fundamentals are still assessed, within the new AI-assisted process, and that you must be able to explain any code you submit.

Which AI tool should I use for a Canva interview?

Canva names Copilot, Cursor and Claude as examples. Use the one you are most fluent with, and check the instructions you are sent for your interview.

Keep going

Start the practice roundAll AI-assisted rounds

Sources

  1. Canva engineering blog: "Yes, you can use AI in our interviews. In fact, we insist"
  2. The Register: Canva now requires candidates to use AI coding assistants in interviews (June 2025)

Placed is not affiliated with Canva. Company names are used to describe publicly reported interview formats.