Google is piloting an interview format that lets software candidates use an approved AI assistant during code comprehension rounds.
Google is testing a major shift in how it hires software engineers: letting candidates use an AI assistant during parts of the interview.
Misryoum reports that the change is part of a broader overhaul, designed to better reflect how software development is evolving.. In the pilot. candidates for junior to mid-level roles will be allowed to use an “approved” AI assistant during a “code comprehension” round. beginning in the second half of the year.
This matters because interviews are often used to measure more than coding ability. When AI is brought into the process, companies are effectively testing a new blend of skills: how candidates work alongside tools rather than only how they perform unaided.
Under the proposed format, candidates are expected to read, debug, and optimize code drawn from an existing code database.. Interviewers would assess “AI fluency. ” including prompt engineering. output validation. and debugging skills. according to details in the internal document reviewed by Misryoum.
Misryoum also notes that the scope of the pilot would start with selected teams in the US, with plans to scale to a wider set of teams and regions later if the trial performs well. During the pilot, Google’s own AI model, Gemini, would be the assistant made available to candidates.
The key business angle here is fairness and signal: companies want to understand whether candidates can use AI responsibly and effectively, especially in tasks that mirror day-to-day engineering work.
Beyond the code comprehension round, Misryoum reports that Google intends to test other modifications to its interview structure.. One planned adjustment would add a technical design discussion to the “Googleyness and Leadership” round. which has typically leaned more toward behavioral assessment.. For more junior candidates. Misryoum says one technical round would be replaced with an interview focused on tackling open-ended engineering challenges.
Misryoum adds that early pilots would be rolled out across multiple organization areas. including Cloud and its platforms and devices unit. with some testing beginning this month.. The move also reflects a broader trend in tech hiring. where some firms have already experimented with allowing AI tools during technical evaluations.
At the end of the day, this kind of policy shift could reshape how candidates prepare for interviews, placing greater weight on verification, debugging, and tool-driven problem solving rather than purely memorized approaches.