Bots vs. Bots: What Happens When Candidates and Employers Both Use AI

TL;DR: AI is officially on both sides of the interview table.

Candidates are using AI to find jobs, improve resumes, prepare for interviews, and even generate answers in real time. Employers are using it to manage growing application volume, screen talent, detect fraud, and make hiring more efficient.

The challenge now isn’t whether AI belongs in hiring. It’s figuring out how to use it without losing trust...and potentially losing out on great candidates along the way.

In practice, that means:

• Getting clearer about where candidate AI use crosses the line from assistance to misrepresentation. (Check out this case study to see how one team handles this).
• Being transparent about how employers use AI and where humans remain responsible.
• Looking beyond increasingly polished resumes to validate real skills.
• Adding safeguards like identity verification without creating unnecessary friction.

AI used to be mostly an employer-side conversation.

What should recruiters automate? Where could AI save time? How could technology help teams make sense of more candidates, more data, and more complexity?

Then candidates got access to many of the same tools.

So what happens when AI shows up on both sides of the interview table?

That was the focus of our recent Bots vs. Bots: What Happens When Candidates and Employers Both Use AI webinar. Employ’s Sara O’Donnal and Dara Brenner joined Amanda Knight (Associate VP, Human Resources at American Specialty Health) and Taylor Liggett (Chief Growth Officer at ID.me) to unpack findings from Employ’s 2026 Job Seeker Nation Report and discuss what they mean for hiring teams today.

And based on the questions coming from the audience, this isn’t a hypothetical problem anymore. Recruiters are already encountering resumes that perfectly mirror job descriptions, unusually polished interview answers, questions about candidate identity, and growing uncertainty about which hiring signals they can still trust.

If you missed the live webinar or you’re just looking for a little refresh on the conversation, we’ve got you covered. Here are the biggest takeaways from the webinar—and a few insights you can bring back to your own team. 

What Happens When AI is On Both Sides of the Interview Table

Candidate AI Use Isn’t the Problem. Knowing Where to Draw the Line Is.

Despite all the attention around AI-generated applications, most candidates still aren’t using AI during their job search.

According to the 2026 Job Seeker Nation, only 28% say they do.

Among those who are, the most common uses are fairly practical: 53% use AI to find or match with jobs, 47% to write or review resumes, 38% to draft cover letters, and 36% to generate interview questions.

For the most part, AI is helping candidates find opportunities, improve how they present their experience, and prepare more efficiently.

But one finding made the conversation much more complicated:

37% of candidates who use AI say they use it to get suggested interview responses in real time.

And that’s where employers have to start making judgment calls.

If using AI will be part of someone’s actual job, asking candidates to avoid it entirely during hiring may not always make sense. But there’s still a meaningful difference between using AI to communicate your experience more effectively and using it to represent knowledge or skills you don’t actually have.

That distinction came through clearly in the webinar discussion. Attendees described candidates repeating interview questions before giving textbook-perfect answers, struggling to connect polished responses to specific experiences, and submitting resumes that seemed almost too closely tailored to the job description.

But the audience raised an equally important counterpoint: some of those same behaviors can have perfectly legitimate explanations. A candidate repeating a question may be practicing active listening. A pause may mean they’re thinking. A highly polished application may simply belong to a highly prepared candidate.

Which is why trying to identify every possible “AI tell” isn’t a particularly reliable hiring strategy.

A better approach is to make the conversation harder to fake.

Instead of stopping at a general behavioral question, go deeper. For example: If a candidate brings up a successful past project, ask what aspects they personally owned, why they made a particular decision, what happened as a result, and what they would change if they faced the same situation again.

AI can generate a polished general response but normally falls apart once you go a layer deeper. Genuine experience tends to hold up...even when the follow-up questions start.

Candidates Want Rules for Employer AI Use, Too

Candidates may be using more AI themselves, but they’re also paying close attention to how employers use it.

But here’s what might surprise most TA pros; they’re not necessarily opposed to it.

Sixty-three percent of candidates are comfortable with AI being used during hiring, particularly in supportive parts of the process. Forty-one percent are comfortable with AI recommending relevant jobs, 40% with it reviewing a resume or application, and 38% with it sending application updates.

What they don’t want is to be left guessing about what technology is doing behind the scenes.

Ninety-one percent believe employers should be transparent about when and how AI is being used.

That was one of the clearest themes of the webinar. Candidates aren’t simply asking employers to disclose that AI exists. They want to understand what role it plays, where human judgment comes in, and who is ultimately responsible for the outcome.

Dara shared a simple example in the live chat:

“AI helps us organize application data, but human recruiters make all hiring decisions.”

That works because it explains more than just the presence of AI. It explains its boundaries.

And those boundaries matter. 

During the live session, audience members shared their opposing views on what some might consider one of the more mainstream and embedded AI uses cases: AI interview notetakers. Some attendees saw value in using them to stay more present in the conversation rather than splitting their attention between the candidate and their notes. While others raised questions about recordings, transcripts, privacy, and sensitive information candidates may share during an interview.

But these concerns aren’t limited to one tool, and in fact, they reinforce the broader point: transparency needs to go beyond simply telling candidates that AI is being used. Candidates should clearly understand what information is being captured, why it’s being collected, and what role that technology plays in the hiring process.

That’s what makes the difference between a hiring process that breaks down trust and one that builds it up. 

How Teams Are Build Trust in an AI-Powered Hiring Process

Balancing Security and Suspicion

During the webinar, hiring teams shared examples of candidates listing certifications they couldn’t verify, resumes that appeared to copy job descriptions word for word, and even situations where one person interviewed but someone else appeared later in the process.

But they aren’t the only ones experiencing fraud and scams. Candidates are experiencing fraud from the opposite direction, too.

Fifty-three percent of job seekers say they’ve encountered a job posting they believed was a scam. Their biggest warning signs included jobs that seem too good to be true, requests for payment, and requests for personal information too early in the hiring process.

That creates a trust paradox around identity verification.

Employers may ask candidates to verify their identity because they want to make the hiring process safer. But candidates have also been taught that giving personal information to an unfamiliar employer can be a warning sign of danger to come.

The live Q&A reflected that uncertainty. Hiring teams wanted to know when verification should happen, what information they should request, and how organizations should think about verifying remote candidates.

There’s unlikely to be one universal answer. 

But more important than timing is creating transparency around the process. If identity verification is part of the process, explain why it’s needed, when it will occur, what the candidate will be asked to provide, and how that information will be handled.

Because adding a safeguard doesn’t automatically make a hiring experience feel safer. The way it’s introduced? Now that can make all the difference.

Screening for the Right Signals

The average number of applications per job increased from 207.2 in 2024 to 257.5 in 2025. That’s roughly 50 more applications per opening.

Which means, for recruiting teams, manual reviews are quickly becoming more unsustainable. And because of that, they’re turning to an AI-powered answer.

AI-powered screening can help organize candidate information, identify relevant qualifications, and give recruiters more time to focus on the people who deserve a closer look.

But candidates see that automation happening, too.

Thirty-four percent believe they’ve already been automatically rejected by AI.

Whether an algorithm actually made the decision in every case almost becomes secondary. If candidates believe nobody ever looked at them as a person, trust in the process takes a hit.

That’s why human oversight has to be more than a behind-the-scenes policy. Candidates should have a reasonable understanding of where AI is helping and where people remain accountable.

At the same time, employers need to be careful about replacing one imperfect screening signal with another.

That became apparent when the audience discussed whether things like a missing LinkedIn profile should be considered a fraud signal. Recruiters working in government and other industries quickly pointed out that qualified candidates may have little reason to maintain a detailed LinkedIn presence. And fake profiles can be created too.

The lesson is simple: no single signal tells the whole story.

As AI makes it easier to create polished applications, hiring teams need a fuller picture of candidates, not just a longer list of things that look suspicious.

Looking Beyond the Resume

When nearly anyone can use technology to create a polished resume, employers have more reason to focus on what candidates can actually demonstrate.

And candidates seem to agree.

Seventy-three percent say they’ve completed a skills assessment or work sample during a hiring process, and 79% agree skills assessments can better help employers determine whether someone can perform the job.

They also want employers looking beyond work history. Outside of experience, candidates say they want to be evaluated on communication skills, their potential to learn and grow, problem-solving ability, and demonstrated skills.

But the webinar audience also surfaced the tradeoff.

How do you know someone isn’t using AI to complete an unproctored assessment? Would a timed exercise give you a better signal? And at what point does asking candidates to produce work for free create more friction than value?

Those questions matter because the answer to AI-polished resumes can’t simply be adding more hoops to the hiring process.

A useful assessment should measure something meaningful for the role and ask for a reasonable investment from the candidate. That might mean a short, structured exercise, a deeper experience-based interview, or a work sample later in the process once both sides have established real interest.

The goal isn’t more validation for validation’s sake. It’s stronger evidence of whether someone can actually do the work.

Trust Has to Be Built Into the Process

It’s clear hiring teams are responding to very real challenges. Applicant volume is rising. Fraud is harder to spot. Resumes are easier to optimize. And recruiters still need ways to identify the people most likely to succeed.

Technology can help solve those problems. But every new tool also changes the candidate experience. And creates more questions: 

• If AI screens my application, who makes the final decision?
• If you verify my identity, why do you need that information?
• If you record an interview, what happens to the transcript?
• If you ask me to complete an assessment, what are you actually evaluating?
• And if you’re concerned about my use of AI, where exactly is the line?

That’s why trust can no longer be treated as something that sits alongside the hiring process. It has to be designed into it.

That doesn’t mean avoiding AI or adding a human touchpoint simply for the sake of having one. It means being deliberate about what technology does well, where human judgment matters, and how clearly candidates understand the difference.

As AI becomes more capable on both sides of the interview table, those decisions will only become more important.

Because the biggest takeaway from Bots vs. Bots wasn’t that employers need to win an AI arms race with candidates. It was that hiring teams need better ways to understand what’s real, communicate what’s happening, and make thoughtful decisions with the technology now available to both sides.

And the teams that get that balance right will be better positioned to use AI without losing the trust that good hiring still depends on.

Want to explore the data behind the conversation? Dive into the 2026 Job Seeker Nation Report for more insights from 1,500+ workers on how AI, trust, and candidate expectations are changing the job search.