How do you use AI tools when you code?
Interviewers now ask this often. They want to know whether you can use AI productively and responsibly.
How would you answer? Think about what you use it for, how you verify its output, and where you don't use it.
Reveal model answer
Model answer
"I use AI assistants every day, mostly as a fast pair programmer. I use them for:
- First drafts and boilerplate: forms, tests, API calls
- Explaining unfamiliar code, libraries or error messages
- Exploring options: 'what are two other ways to approach this?'
But I own everything I commit. I review AI code like a teammate's pull request: check edge cases, run the tests, and make sure I can explain every line. For example, an assistant once gave me a pagination function with an off-by-one error that only showed up on the last page. My tests caught it.
I'm careful with context and privacy: I never paste secrets or customer data, and I follow the company's AI policy.
And I keep my fundamentals sharp, because they're what let me tell when the AI is wrong."
Why this works: it's specific, it shows judgment, it gives a concrete example, and it reassures the interviewer that you add value beyond the tool.