People using AI to generate their answers in real time during interviews

[wattnext]

I've been interviewing for a professional role this week and I strongly suspect that at least two of the candidates were using AI to listen to the questions and generate answers in real time, which they then read out. It was like interviewing robots. The answers technically met the criteria very well, but were overly long and detailed. Their delivery was very robotic ... no smiles, no natural conversational pauses, no natural inflection to make it engaging ... perhaps future iterations of the tech might remind candidates about those things too! 😱

They didn't get the job, and were given carefully drafted feedback!

I think given how much AI is used in the recruitment process it wouldn't surprise me if some candidates genuinely think that AI is also acceptable for them to use in this situation.

I'm sure we had this with one candidate last week.

We decided to prioritise in person interviews for our main round going forward. The roles are hybrid anyway.

Think returning to in person recruitment will also give a better understanding of candidate fit to our team.

I think it's different to use AI as a time saver (to sift through lots of applications) than to use it to mis-represent who you are.

Don't get me wrong, I wouldn't do this or recommend it. I do think though that when anything starts to be normalised then people will push the boundaries on what's acceptable.

Make it clear from the outset that use will mean you will not be offered the job.

Yep, that's the plan. Though it's got to be communicated carefully as it's difficult to prove someone is using it.

Each of the answers started with a deliberate pause ... "that is a very good question" or "thank you for that question" ... to give the AI time to generate a response. The answers were textbook examples of the "STAR" technique, but unnecessarily long and detailed, and they all ended abruptly with a perfunctory "Thanks".

All interviews in my organisation start with a statement underlining the fact that AI is not to be used and if it is found to have been it will be seen as Gross Misconduct (for present employees) or will mean they will be ineligible in future. But some roles are having face to face interviews (a problem as we are widely distributed and it takes a lot of time).

Last round of interviews I was involved in, the candidates were given the questions in advance so some definitely had AI answers, it just meant that follow up questions were more important, asking them about their initial answer. It was soon apprent who knew their stuff and who didn't.

At the last day of interviews I did, we had extra information in our briefing about how to spot potential AI answers and what to do about it. I don't think any of the candidates I saw used it, but there were some questions from other panels. Going back to F2F interviews would help with this (and I prefer it to online anyway!) but the costs/practicalities mean this is very unlikely.