Ask a candidate "So, what are your thoughts on AI?" and most hiring managers think they are making small talk. They are not. This one question is quietly becoming one of the most revealing filters in a senior interview, and the answer tells you far more about how a person works than any line on their resume. After 19 years placing senior and executive talent across cybersecurity, enterprise software, AI infrastructure and data centers in the DACH region and the Nordics, I have watched this question decide final rounds. Here is what a good answer actually reveals, and how to read the room when you ask it.

Why the AI interview question tells you more than an opinion

The question sounds harmless, which is exactly why it works. A candidate hears "what are your thoughts" and assumes you want an opinion. So they give you one. "I think AI is really exciting." "It's going to change everything." "I'm not too worried, my role still needs real human thinking." None of those answers is wrong. They are just empty. An opinion is not evidence, and as a hiring manager you already know the difference from every deal review you have ever sat through.

You are not asking to start a debate about the future of humanity. You are checking four things at once: is this person adaptable, are they curious, are they actually using these tools or just talking about them, and do they have enough judgement to know when to trust the machine and when to override it. The candidates who understand that they are being tested on how they work, not what they believe, are the ones worth building a team around for the next three years.

What a weak answer sounds like

Two patterns should make you lean back in your chair. The first is defensiveness. You say the word AI, and the candidate hears an accusation. They start justifying themselves: "Well, I do use ChatGPT for my applications, but I always check the work." Nobody asked about their resume. When a candidate answers a question you did not ask, you are watching how they behave under mild pressure, and it is not reassuring.

The second pattern is more common and easier to miss because it sounds positive. The candidate who tells you "I use AI for everything" thinks they are signalling that they are modern and fast. To a trained ear it signals the opposite: no boundaries, no quality control, and a person who might drop confidential customer data into a public tool without a second thought. In a boardroom where the real conversation is about how to make teams AI-ready without breaking data security or customer trust, unlimited enthusiasm is a risk flag, not a selling point.

The three-part answer strong candidates give: position, proof, limitation

The best answers you will hear follow the same three-part shape, whether or not the candidate knows it has a name. Learn to listen for it, because it separates people who use AI with judgement from people who are performing familiarity.

Position is the calm opening. "I see AI as a tool, not a threat. Used well, it makes good people more effective." No panic, no overselling. It tells you they have actually thought about where this fits.

Proof is where most candidates fall apart, and where the strong ones separate themselves. You want a concrete example from their own work, ideally with a number. In a sales context that sounds like: "I use AI to speed up account research before a first meeting, to structure call notes, to prepare clean follow-ups. My weekly reporting used to take three hours, now it takes forty minutes." That is the moment the answer flips from talking about AI to showing it. If a candidate cannot give you a specific example, they are probably not using the tools the way they claim.

Limitation is the part that sounds like weakness and is actually strength. "Of course I sense-check everything, AI is only as good as the person directing it." Or: "Customer data and confidential deal information never go into a tool unfiltered." A candidate who names the risk without being prompted has the judgement you cannot train quickly. The most impressive answer is never "AI is amazing and I use it for everything." It is "I understand the value, and I understand the risk."

How to read the answer by seniority

Calibrate your expectations to the level you are hiring for. From an early-career seller, honesty beats polish. Someone who says they are building AI into their preparation right now, into research and structuring their arguments, and who understands the limits, is telling you the truth. That is fine. Pretending to have it all figured out at that level should worry you more than admitting they are still learning.

From a senior AE or a sales leader, the bar moves. Personal tool use is table stakes. What you are really listening for is whether this person can bring a team along on AI without quality or customer trust slipping. Same question, very different answer. If a candidate for a leadership role only talks about their own productivity and never mentions the team, you have learned something useful about how they will actually lead.

A final round that turned on this one question

A concrete example. I recently had a very strong account executive in a process with a cybersecurity vendor, final round, DACH. Clean technically, solid numbers, easy to like in the room. In the last conversation, the VP of Sales asked exactly that: "So, what are your thoughts on AI in sales?" The candidate, nervous like everyone in a final round, slid straight into the defensive trap. He spent a full minute explaining that he had used ChatGPT for his application materials but of course checked everything himself. He defended something nobody had asked about.

In a parallel process, another candidate got the same question. She answered with a question of her own: "Can I quickly check, do you mean how I use AI in my pipeline work today, or how I think it's changing enterprise sales?" The VP clarified. She answered in three sentences: position, one concrete example from her account preparation, and a clear limitation around customer data. The VP called me afterwards, and his line stuck with me: "The second one didn't talk about AI, she showed me how she thinks." She got the offer. The tools were never the point. Judgement was.

What to listen for as a hiring manager

When you ask the AI question, stop grading the opinion and start grading the structure. Did they clarify before answering, or rush in? Did they give you proof from their real work, or float on generalities? Did they name a limitation on their own, or leave you wondering whether they have any boundaries at all? A candidate who clarifies, proves, and self-limits is showing you the exact behaviour you want when they are sitting across from your most important client. The question is small. What it reveals is not.

FAQ

What does the AI interview question actually test?

It tests how a candidate works, not what they believe. Specifically: adaptability, curiosity, whether they genuinely use AI tools, and whether they have the judgement to know when to trust a tool and when to override it. The opinion itself is almost irrelevant.

What is a red flag in a candidate's answer about AI?

Two. Defensiveness, where the candidate justifies using AI on their application when nobody asked, and unlimited enthusiasm, where "I use AI for everything" signals no boundaries and a risk to data security and customer trust. Both suggest weak judgement under pressure.

How should the answer differ between junior and senior candidates?

From junior sellers, expect honesty about learning and clear awareness of limits. From senior AEs and sales leaders, expect more: evidence that they can bring a whole team along on AI without letting quality or customer trust slip. A leader who only talks about personal productivity has answered the wrong question.

Hiring for judgement in an AI-heavy market

The candidates worth hiring are the ones who keep ownership in a world full of tools. If you are building a senior sales or leadership team in cybersecurity, enterprise software, AI infrastructure or data centers across DACH or the Nordics, that filter matters more every quarter. If you want a sharper read on how to assess it in your own processes, reach out on LinkedIn or at nordh.de.