Interviewing in the age of AI: How to see past the scripts and find the real candidate

When AI shows up uninvited 

Job interviews used to be about conversations between humans – a way for the hiring team to go beyond what was listed in a candidate’s CV and get to know them better as a person. But those face-to-face interactions have become increasingly rare. The majority of jobseekers (54%) told us that they’ve participated in an AI-led interview.

Because they’re more likely to interview with an AI tool, candidates feel more pressure to appear polished and present scripted answers. As a result, they’re also turning to AI tools to help. This situation leads to what Greenhouse CEO Daniel Chait calls a “doom loop”. While strapped TA teams try to boost efficiency and candidates aim to present themselves in the best possible light, this overreliance on AI tools is making everyone miserable. 

Both sides are saying, ‘This is impossible, it’s not working, it’s getting worse.’

^{– Daniel Chait, Greenhouse CEO}

So how do you escape? You may not be able to ban AI completely, but you can interview better. Let’s consider what that might look like.
‍

Why traditional interview signals are harder to read 

Before AI was widely available, interviewers looked to candidates’ preparedness and confidence as signals of competence. But that approach no longer works. Polish is now nearly as likely to come from an artificial source as from lived experience. New Greenhouse data shows one-third of recruiters (35%) say they’re seeing AI used live during interviews.

With less variation in candidates’ responses and more surface-level sameness, it’s become harder to spot quality talent. You have to dig deeper to find authentic answers and genuine insights into candidates.

‍

Structure is the antidote to complexity

So how exactly do you get candidates to go beyond surface level in interviews? It all starts with structured interviews that promote consistency in your process.

Creating a list of standardised questions that you’ll ask all candidates for a given role makes it easier to compare candidates fairly. And using interview scorecards to keep track of interviewers’ assessments anchors evaluations in evidence rather than vibes. Plus, relying on a consistent rubric will give you data points that limit the role of subjectivity and bias in your decision-making.

Here are a few additional pointers to help you ace interviewing in the age of AI.
‍

Ask questions that surface lived experience rather than memorised talking points 

One of the best ways to avoid overly polished answers that can be easily generated by AI is to surface a candidate’s lived experience. Greenhouse Talent Acquisition Manager Bronté Chappell offered the following advice: “Prioritise behavioural and situational questions that require candidates to share what they did, how they made decisions and what they learned along the way. You’ll uncover real experience, self-awareness and learning agility – qualities AI tools can’t really fabricate.” 

Wondering what exactly you should ask? Try prompts like the ones we shared in this interview questions template:

• “Tell me about a time when…”
• “What role did you play in solving the issue?”
• “What would you do differently now?”

These types of questions work because AI struggles with authentic details. One telltale sign that a candidate is describing their own lived experience? Real examples include nuance and trade-offs. By doing this type of detective work, you’ll be able to surface depth, judgement, decision-making and a growth curve.
‍

Focus on outcomes and evidence, not only delivery 

It’s easy to equate confidence with ability – especially for roles like sales that require strong communication skills. But interviewing in the age of AI changes the game. “A confident delivery is helpful, but it’s not the whole story,” said Bronté. 

To go deeper into a candidate’s responses, Bronté suggested using scorecards to look for evidence of skill, role readiness and impact a candidate has demonstrated in past work.

Strong communication can elevate a response, but outcomes and examples should carry the weight.

^{– Bronté Chappell, Greenhouse Talent Acquisition Manager}

As you’re listening to a candidate’s answers, try to identify proof of measurable impact, clarity of role scope, stakeholder relationships and lessons learned after failure. And don’t be afraid to ask follow-up questions to gain more clarity!

‍

Watch how candidates collaborate in the conversation

Interviews are a little different from other types of conversations you have in the workplace, but they can still provide helpful insights into a candidate’s communication style and personality. “It’s important to understand how someone engages: asking clarifying questions, responding thoughtfully to follow-ups, showing curiosity and building on ideas,” said Bronté. “These signals help you understand how they might partner with hiring managers, peers and cross-functional teams – not just whether they interview well.” 

Looking out for collaboration signals is important because it’s rare for anyone to work in complete isolation these days. Clear communication is a prerequisite for nearly any knowledge worker role. 

There’s another benefit to keeping a close eye on collaboration skills: AI can’t fake these dynamics. Real-time interaction requires adaptability – not simply sticking to a script. Following up appropriately takes situational awareness. And good teamwork is visible in these micro-moments.

‍

Make AI expectations clear to candidates 

One of the biggest challenges for candidates is knowing what’s acceptable and what isn’t when it comes to AI use – 27% of candidates told us they’ve never seen an employer policy on AI use in hiring. You can take a proactive approach to setting expectations and being transparent. This takes the guesswork out of interviewing and helps eliminate some of that doom loop dynamic we mentioned earlier.

Here at Greenhouse, for example, we published our own guidelines for using AI in our interviewing process. You can take our guidelines as inspiration or write your own from scratch. Either way, we recommend the following:

• Explain what’s welcome (e.g., candidates can use AI for research or interview prep)
• Explain what’s off limits (e.g., candidates should not use AI to generate real-time responses during interviews)
• Clarify how your TA team is using AI (e.g., disclose where in the process you’re using AI tools, such as for sourcing or scheduling, and where you’re not, such as when you make final hiring decisions)
  ‍

Human judgement remains your edge

One thing is clear: AI isn’t going anywhere. Instead of trying to ignore this reality, the best recruiters are adapting with structure and nuance. 

A good rule of thumb is to use AI for logistics and preparation, while leaning as much as possible on individuality during in-person interactions. Don’t be afraid to remind candidates that you want to hear about their lived experience and that their mistakes and learnings are so much more powerful than seemingly perfect but ultimately empty answers.

When in doubt, bring it back to structure and consistency. The more you ask the same questions of every candidate, the easier it’ll be to hear the answers – and confidently select the candidates – that genuinely stand out.
‍

Join this upcoming webinar for an exclusive look at findings from the Greenhouse 2025 AI and Hiring Survey, and learn how you can restore trust and authenticity in an AI-driven hiring landscape.