Open your applicant tracking system right now and you'll probably notice something: the resumes are starting to sound the same, same structure, same confident bullet points, same phrase about being a "results-driven professional." That's not a coincidence: it's AI-generated applications, and they're arriving faster than most hiring teams can even read them.

This isn't a small trend. Candidates are using AI to write resumes, tailor cover letters to every job description in seconds, and in some cases, generate answers in real time during interviews. Meanwhile, recruiters are the ones absorbing the impact, sorting through a flood of applications that all look qualified on paper and sound polished in person, with less signal than ever about who's actually behind them.

The instinct is to fight AI with more filtering, but that's the wrong move. The better one is changing what you're actually testing for.

Why Polish Isn't the Problem

A well-written application was never proof of anything beyond writing ability, and now it's proof of even less. SHRM's recent guidance on this points hiring teams toward structured interviews, the same core questions, asked the same way, to every candidate for a role, scored against the same criteria, because structured formats aren't just fairer, they're harder to game. They're built around how someone thinks through something in the moment, not how well they polished an answer in advance.

Behavioral questions do similar work. Instead of asking a candidate to describe their strengths, ask them to walk through a specific situation: a project that went sideways, a decision they'd make differently now, a time they had to change someone's mind. AI can help someone draft an answer to "what's your greatest weakness," but it's a lot less useful mid-conversation, once you ask a genuine follow-up question that only makes sense if they actually lived the story they just told.

The Gap Isn't the Candidates, It's the Process

Here's the harder truth for a lot of hiring teams: the barrier usually isn't candidates gaming the system, it's that the process was never built to hold up under it. One recent industry finding put a number on it: 42% of HR professionals say their biggest obstacle to using AI well in recruiting is that they don't yet know how to use it effectively themselves, on either side of the table. So teams end up reacting to AI-written applications with the same unstructured interviews and gut-feel scoring they used before, and wondering why it's gotten harder to tell candidates apart.

The fix isn't banning AI from the hiring process, it's being honest about what you're actually evaluating. Let candidates use it to get organized, just stop grading them on the writing and start grading them on the thinking. That means:

• Standardize your questions so every candidate is measured against the same bar.

• Ask for specifics, not summaries, since real experiences hold up under follow-up questions and fabricated ones don't.

• Score consistently, using the same rubric across every interviewer, so "polish" stops quietly outweighing substance.

The talent market hasn't gotten worse, it's gotten louder, and the hiring teams who adjust their process, not just their scrutiny, are the ones who'll still be able to tell who they're actually hiring.

Phizenix's Talent Solutions team builds hiring processes designed for exactly this moment, structured, fair, and built to find the signal underneath the noise. If your interview process hasn't caught up to how candidates are actually applying, let's talk.