Are You Interviewing a Candidate—or Their AI?

You’re a few questions into the job interview and something feels off. The candidate’s answers are polished. (Maybe too polished.) There’s a slight delay before each response. And their eyes keep drifting off camera. You’re pretty sure ChatGPT is feeding them answers. Welcome to hiring in the age of AI, where technology doesn’t just help candidates prepare—it can also do the interview for them. This might feel unsettling, but it’s not surprising, according to Mike Kyle, managing director of technology recruitment at Coda Search & Staffing, an executive recruiting firm. “If you use ChatGPT to write job descriptions and AI to screen applicants, you shouldn’t be too upset when candidates use it in interviews.” At a time when AI is inevitably in the mix, leaders face a new challenge: how to evaluate someone’s true abilities when a chatbot might be doing the talking. Hiring managers might be tempted to ask candidates to share their screen to prove they’re not using AI. But that can backfire because it signals distrust and can turn off candidates. The smarter approach is to rethink how you conduct the interview, says Tiziana Casciaro, a professor of organizational behavior at the Rotman School of Management at the University of Toronto. “I would not go into detective mode and try to police what the candidate is doing,” she says. “Your goal should be to create an interview environment where just reading from a ChatGPT script won’t work.” In other words: Don’t worry about catching AI use; instead, focus on whether candidates have the human skills AI can’t replicate. Here are five to keep in mind. How well does the candidate read the room and respond to social cues? Sure, there are tells: awkward pauses, wandering eyes, stilted delivery. But whether that’s AI or interview jitters, what matters is whether candidates have core interpersonal skills, like the ability to read social cues and adapt. Kyle recommends chatting with candidates before diving into the heart of the interview to establish a baseline for their speech patterns. “Get a sense of their communication style.” Once you have that baseline, watch how they handle shifts in the conversation. Change your tone or body language, gently push back on an answer, and see if they notice and adjust. That responsiveness tells you about their self-awareness and whether they can read people. “Emotional intelligence, or soft skills, are not separate from competence,” says Casciaro. “These skills determine how candidates will work with your team, your colleagues, and your customers.” How do they reason through complex ideas and connect information? Even if a candidate is using a chatbot during the interview, there’s a big difference between relying on it entirely and using it to help shape their own thinking, says Casciaro. In her work as a professor, she notes that even when students have access to the same AI tools for written assignments, some stand out. “You can tell who’s thinking critically by the logic in their argument and how they connect the dots,” she says. “Some build on information and refine their ideas, while others simply parrot what they’ve been given.” Kyle recommends hiring managers look for similar qualities. He suggests giving messy, ambiguous scenarios where there isn’t a clear right answer to see how candidates work through uncertainty. Then, listen for how they reason and connect ideas. “Don’t ask, ‘What do you know about X?’ Ask, ‘Walk me through how you’d handle this situation,’” he says. “I’m more concerned with how a person thinks than what they can recall—we all use technology to help us remember things.” How do the candidate’s questions help move the conversation forward? Another way to determine whether candidates are curious and capable of critical thinking is by seeing how willing they are to ask questions and how well they identify gaps in information. Kyle recommends posing questions that are deliberately vague or missing key details. You want to see if they recognize what’s missing and ask for more information. For example, you might ask a data engineer, “How would you build a data pipeline that moves from A to B?” or a finance candidate, “How would you approach month-end close for our business?” “Those are generic questions. If they give you generic, canned answers they’re just reading a script,” he says. “But if they ask for clarification and dig into details, it shows they’re thinking critically.” How do they respond when the discussion shifts in an unexpected direction? Pay attention to how the candidate responds when you go off script. If they’re smooth on familiar ground but stumble in unfamiliar territory, it shows how they handle unexpected challenges, says Kyle. He suggests surprising candidates mid-interview with a request: Share your screen and walk through something you’ve worked on, such as a project draft, a piece of code, or whatever’s relevant. “This approach might not fit every role,” he says, “but for some positions, it’s a way for candidates to talk through their process, the decisions they made, and what they might do differently next time.” You’re not trying to monitor their tabs, of course. You’re watching whether they stay composed and can think on their feet and explain their reasoning without preparation. Kyle also recommends asking candidates how often they use AI and what they use it for. This helps you see how comfortable they are with the tools, how they apply them personally and professionally, and whether they could help drive adoption within the organization. “In my experience, people who use it regularly tend not to over-rely on it because they know its limits and what it can and can’t do,” he says. How well does the candidate work with others and solve problems on the spot? Some organizations are requiring candidates to attend final-stage interviews in person, notes Kyle. “This makes sense for companies with hybrid schedules, where employees need to be on-site part of the week or month,” he says. “In-person interviews, especially those that involve whiteboarding or case discussions, have the added benefit of being more collaborative.” Still, they can be expensive, logistically complicated, and aren’t always practical, says Casciaro. A simpler alternative is group interviews, where multiple candidates join a video call together, three or four at a time, for a team-based exercise or problem-solving activity. Group interviews have a faster tempo and are more dynamic than one-on-one conversations, she says. They also give a clearer picture of the candidate’s competence and teamwork abilities. You’ll see who listens and builds on others’ ideas, who brings quieter people into the conversation, and who dominates or talks over everyone else. “Yes, candidates need technical skills. But if someone can bring people together, manage complicated interactions, and make sense of multiple perspectives, that matters enormously.” . . . Chances are, candidates are going to use AI in your job interviews, and even if you notice it, there’s only so much you can do about it. Rather than trying to control every detail, focus on what you can learn about them: how they handle ambiguity, reason through problems, and adapt to the unexpected. After all, “If your interview can be passed by a chatbot, that’s not a candidate issue,” says Kyle. “That’s your problem.”

Are You Interviewing a Candidate—or Their AI?

Summary.

Partner Center