Google lets software engineers use AI assistants in job interviews

Google is piloting a new interview process for software engineering candidates that will let them use an AI assistant during the "code comprehension" round, according to an internal document reviewed by Business Insider [1]. The change is part of a broader overhaul of Google's interview process, which the document says is being made "to better align with the modern engineering landscape." Starting in the second half of the year, Google will permit the use of an "approved" AI assistant, with Gemini, the company's own AI model, being used by candidates during the pilot phase.

The move reflects the significant changes that AI has brought to software developer roles. In late 2025, Anthropic and OpenAI launched new models that dramatically improved the capabilities of coding agents [1]. Now, three-quarters of new code created inside Google is generated by AI, the company said in April [1]. This shift is happening elsewhere as well, with OpenAI president Greg Brockman recently saying that AI has gone from writing 20% of code to 80% [2]. The interview process hasn't kept up with these changes, with some engineers feeling that traditional tests no longer reflect "what their job will actually look like," according to Stefan Mai, a former engineer for Meta and Amazon [2].

Google's move follows what some tech startups have been embracing for a while. Graphic design giant Canva and AI coding startup Cognition are among the companies allowing candidates to use AI in technical interviews [1]. Emily Cohen, who heads people and operations at AI coding company Cognition, told Business Insider that it has changed its interview process to incorporate AI use, saying "I guess this is like asking a kid to take a math test without a calculator" [1]. The new interview process is designed to be "human-led, AI-assisted" and will better simulate a software engineer's "workflow in the GenAl era," according to the Google document [1].

The shift to AI-assisted interviews is part of a broader trend in the tech industry, with companies starting to focus more on questions that show how candidates think through issues and weigh trade-offs rather than raw coding capabilities [2]. However, it's still an "unsolved problem," according to Mai, with companies struggling to figure out how to evaluate candidates who are increasingly using AI to do their jobs [2]. As the industry continues to evolve, it's likely that companies will need to adapt their hiring processes to keep up with the changing landscape.