A lot of people ask me if using AI during interviews is allowed, and the honest answer is: it depends. We’re in a weird transition phase where some companies are starting to allow AI in technical interviews, others still treat it like cheating, and many simply haven’t updated their process yet. I mention how Meta has publicly talked about experimenting with interviews where candidates can use AI tools as part of their coding workflow—because in the real world, engineers aren’t coding in isolation anymore. We use copilots, agents, and AI-assisted IDEs daily, so testing you without AI can be testing a workflow that barely exists. But “AI allowed” doesn’t mean “easy.” It becomes a different kind of test: how you think, how you design, and how you collaborate with tools—while the interviewer is still evaluating you, not the model. In the video, I break down how to handle this: start with clarification questions (constraints, edge cases, I/O, performance), structure your solution before writing code (architecture, data structures, algorithm, complexity), then use AI transparently as a scaffold and review/modify the logic. And if something breaks, don’t just prompt “fix this”—diagnose first, explain your reasoning, then guide the AI with a structured prompt. The engineers who stand out are the ones who know what code should exist before AI writes it.