Google pilots Gemini AI in software engineering interviews

Google is rolling out a test that lets select candidates use its Gemini AI model during specific coding interview rounds. The move mirrors how developers now lean on AI tools in everyday programming.

What changes for candidates?

  • Applicants can query Gemini while tackling the code comprehension round.
  • The AI can help spot bugs, suggest fixes, and improve performance, but the interview remains human‑led.
  • Only the designated round allows AI help; other stages stay traditional.

Why Google is making the shift

Google sees AI as a standard productivity aid, not a shortcut. By letting candidates use Gemini, the company aims to evaluate how engineers incorporate AI into real‑world problem solving.

Pilot scope and timeline

The pilot starts later this year with junior and mid‑level roles on U.S. teams, including Cloud and Platforms & Devices. It will run across multiple divisions before a broader rollout.

Other interview updates

  • “Googleyness and Leadership” rounds now require candidates to explain technical architecture behind past projects.
  • Entry‑level coding interviews will replace a strict algorithm test with an open‑ended engineering challenge.
"We’re evolving our interview process to reflect the AI era," said Brian Ong, VP of recruiting at Google.

This pilot marks a larger industry trend: generative AI tools are becoming embedded in software development, and hiring practices are adapting to match.