Google allows Gemini in interviews

- Google is piloting software-engineering interviews where some candidates can use Gemini during coding rounds, a real break from the usual no-tools setup. - The first target is Google’s “code comprehension” round later in 2026, with junior and mid-level U.S. roles on select teams. (businessinsider.com) - The point is simple: Google now wants to test AI fluency alongside coding, because that is already how many engineers work. (businessinsider.com)

Google is changing a very old software-interview rule: you may not have to work alone anymore. The company is piloting interview rounds where candidates can use Gemini while solving coding tasks, starting with some software-engineering roles in the U.S. later in 2026. That matters because th(businessinsider.com)ogle is basically admitting that the job changed, so the test has to change too. (businessinsider.com)ot lets candidates use an approved AI assistant during certain software-engineering interview rounds, and Google confirmed Gemini will be that tool in the first phase. This is not a blanket “use AI anywhere” policy. It is a targeted change inside a broader interview overhaul that Google describes as more reflective of engineering work in the AI era. (businessinsider.com) ### Which interview round gets(businessinsider.com)m from scratch and more about reading existing code, spotting bugs, debugging behavior, and improving performance or efficiency. In other words, it looks a lot more like a normal day inside a large codebase — which is exactly why AI assistance fits there more naturally than in a whiteboard-style puzzle. (thehansindia.com)starts with junior and mid-level software-engineering candidates on select U.S. teams. Reports also point to early testing across parts of Google including Cloud and Platforms & Devices. So this is real, but still narrow — Google is testing the format before deciding whether to expand it more broadly across teams and regions. (msn.com) ### Wha(thehansindia.com) Gemini’s output is correct, and debugging or improving AI-generated suggestions. That is the key shift. The skill being measured is no longer just “can you code alone under pressure,” but “can you use a model without getting fooled by it.” (ajupress.com) ### Why does that matter so much? Because AI coding t(msn.com)e code, explain code, write tests, and debug inside common IDEs — while warning users to validate outputs because the model can still be wrong. The interview change mirrors that exact workflow: use the tool, but verify everything. (developers.google.com) ### Is Google dropping traditional interviews? No — at least not from what is public today. The process is stil(ajupress.com)ugh some are being reshaped. One reported change is that the “Googleyness and Leadership” round will also include discussion of technical design from a candidate’s past projects, and some junior candidates may get a more open-ended engineering challenge instead of one standard technical round. (newsbytesapp.com)ike cheating and more like calculator literacy — useful, expected, but only if you understand the work underneath. That could ripple far beyond Google. If one of the biggest engineering employers starts hiring this way, other companies may feel pressure to test collaboration with AI instead of pretending modern developers never touch it. (businessinsider.com) ### Bott(newsbytesapp.com)eers actually do now — reading messy code, fixing broken things, and using AI carefully rather than pretending the tool does not exist. (businessinsider.com)

Get your own daily briefing

Scout delivers personalized news, insights, and conversations tailored to your role and industry.

Download on the App Store

Shared from Scout - Be the smartest in the room.