Why Tech Interviews Still Assume Nobody Uses AI

I'm working with a trading company that's testing junior engineers on knowledge you can Google in five seconds. Candidates must promise they won't cheat with Google or AI. The rest of the industry isn't far behind. Tech companies are still hiring junior engineers like it was 2023: coding challenges and technical interviews without AI, as if nobody uses it at work.

The Mismatch

Coding challenges and technical interviews without AI assume work happens in isolation. Engineers type code from memory, solve puzzles, and prove they know syntax by heart. In reality, most engineers use Cursor, GitHub Copilot, or similar tools daily. The interview ignores that.

The consequence: you filter for memorization, not real-world problem-solving. You hire people who perform well in an artificial setting. You miss people who would thrive with the tools they actually use on the job.

What Works Instead (Junior Level)

First, interviews should focus on solving practical, real-world problems rather than abstract, theoretical puzzles. Less "implement a binary search tree from scratch," more "here's a messy API response, fix it and explain your reasoning." That resembles actual work.

Second, AI-assisted pair programming. Candidates use AI coding tools (e.g., Cursor, GitHub Copilot) during live coding rounds. Evaluators focus on how well the candidate guides the AI, reviews its code for bugs, and handles hallucinations. That reflects real job conditions. As we've written about why hire an engineer who won't use AI, the calculus has shifted. You want people who multiply themselves with AI, not those who pretend it doesn't exist.

What to Look for in Seniors

For more senior candidates, the bar is different. Can they simplify complex topics? Do they ask clarifying questions before diving in? Can they discuss past technical decisions and their business impact? It's not just coding ability. It's judgment, communication, and the ability to connect engineering work to outcomes.

Grit and Communication

When I'm interviewing candidates, I like to see grit and perseverance, and solid communication skills. Those matter across levels. Technical skill decays if someone gives up when stuck or can't explain what they're doing.

What This Means for You

Interviews that ban AI filter for the wrong signal. Some teams, as we've written about what happens when a CTO makes AI non-negotiable, have already drawn the line. The question isn't whether AI belongs in your hiring process. It's whether your process reflects how work actually happens.

If you're a founder or CTO rethinking how you hire engineers, the moment you stop treating AI as something to police is the moment you start evaluating what actually matters.