I Support Using AI in HR, but Not AI Bots in Job Interviews

This as-told-to essay is based on a conversation with Emily Fenech, a 41-year-old marketing VP based in Nashville. Her identity and employment have been verified by Business Insider. This story has been edited for length and clarity.

I’m a big fan of AI.

I work at AllVoices, a company in the HR space that uses AI to help with some of the manual and employee-relations tasks that HR folks deal with.

Part of my role is checking out AI applications and staying on top of how it’s being leveraged in different use cases. I write about it. I make resources about it. I’m always looking to see what’s out there.

I recently made a LinkedIn post that went viral, which was about the 10 coolest AI applications for HR, like instant guides or reading notes.

A lot of people commented on the post, suggesting different tools, and an AI interview tool kept coming up in the comments. So I decided to run a mock interview with the tool.

That’s when I had a pretty negative experience.

It was robotic and lacked emotional intelligence

When I joined, the voice was robotic. I was looking at a logo, not even an avatar, which I think can also be creepy. But I was just staring into this blankness with a robotic voice asking me high-stakes questions, with my future livelihood on the line.

At first, I thought it could be good as a screener. It created a hypothetical situation of interviewing for an office manager role and asked me to describe my experience in office management.

I told it I had 25 years as an office manager, and it responded with an exaggerated reply that made me feel gaslit. It said something along the lines of, “Wow, that’s so impressive. 25 years of experience.” It then asked me for more details on my responsibilities, and that’s where it kind of fell apart.

Related stories

I’m not an office manager. I’m a marketer. So I said, “I plan birthday parties and order toilet paper,” as a joke. It responded with something like, “Wow, your ability to plan parties is an impressive quality.”

Even though I was using sarcasm and making jokes because I was not qualified for the role, this robot kept telling me how impressive I was. It felt like it would have found something positive to say about anything.

I assumed that this technology was early and no one was actually using it yet, but I was surprised to see comments from people on LinkedIn who said they experienced this or that their company uses the tool.

I think what’s unfair about it is it’s giving you robotic energy. Humans match the energy that they get in a conversation. When that energy was robotic, I felt myself using short sentences and feeling like I didn’t want to talk to it because it wasn’t a person.

AI should stay out of the interview process

I work in the employee-relations space and we use AI for all sorts of things in HR.

For example, I think transcribing conversations is one of the best use cases. I also know a company that uses AI for performance management to keep track of goals by taking inputs from emails and one-on-one meetings with employees. Some employee support AI tools can help employees find their PTO policy or W2 form without needing HR employees to serve as the middleman.

AI is really good for structuring unstructured data, remembering things, and taking notes. But any conversation that requires emotional intelligence, please don’t use AI. It doesn’t get sarcasm and there’s no human cues. It just can’t read the room.

If it can answer a question correctly or give you a suggestion without making a decision, I don’t see the harm.

When it’s deciding from a pool of candidates who proceeds and who doesn’t, it’s obvious to see the potential for harm.