Your CV Screener Is Being Outsmarted - By People Who Aren't Even Cheating

In the first post in this series, we covered the arms race playing out in hiring: candidates hiding invisible instructions inside their CVs, fabricating experience wholesale, and engineering documents to game AI screening software. The conclusion was straightforward — an employer's process needs to be built around more than a single document.

This post is about a different, and in some ways thornier, part of the same problem. Because the arms race isn't just being driven by fraudsters. It's being accelerated by career coaches.

The Three Prompts

A TikTok video from a prominent career influencer recently pulled 7.7K saves. The content? A three-step Claude prompting sequence for optimising your resume before you apply for a job.

The prompts, shown on screen, go like this:

Prompt 1: “Analyse this job description and extract the skills, keywords and success signals that hiring managers are looking for.”

Prompt 2: “Review my current resume against those signals and identify weak points, missing proof and red flags.”

Prompt 3: “Ask me clarifying questions one at a time to improve my resume while telling the truth.”

Read those carefully. This isn’t coaching fraud. The explicit instruction in Prompt 3 — “while telling the truth” — means the AI is being used to surface real experiences the candidate may not have known how to articulate, not to invent ones that don’t exist. Understanding what a job requires, identifying where your genuine background doesn’t yet communicate clearly, and then drawing out authentic examples to fill those gaps: that’s what career coaches have charged $300 an hour for since before the internet existed.

So before going further — this particular piece of advice, taken on its face, is legitimate. Calling it cheating would be unfair to the people following it.

But that’s exactly what makes it worth talking about.

When Good Advice Breaks the System

Here’s the problem no individual candidate caused, but everyone’s behaviour creates together.

When one person uses AI to precisely map their resume language to a job description’s requirements, they gain a genuine edge. When every candidate does it — and 74% of job seekers now use AI somewhere in their application process — the keyword match stops meaning anything. Not because anyone lied, but because the signal is saturated.

AI-powered screening tools, and most applicant tracking systems, are fundamentally pattern-matching engines. They’re looking for keywords, phrases, and experience descriptors that indicate fit. A candidate who uses a sequence like this will mirror the exact language of the job ad back to the screener. They will tick every signal the system was trained to look for. So will the next candidate who watched the same video. And the next.

The tool that was supposed to narrow 500 applications to 50 now produces a shortlist of people who all know how to play it — not necessarily the people who can best do the job.

This is screening inflation. The same dynamic that made keyword-stuffed websites briefly rank well in search engines, before Google learned to look past the surface. The document becomes a game. The game favours the candidates who know the rules, regardless of the underlying reality.

The Real Casualty: The Candidate Who Doesn’t Know the Tricks

There’s a fairness problem underneath this that doesn’t get discussed enough.

A highly capable candidate who hasn’t encountered this kind of coaching, who simply writes their own resume and applies, now competes on unequal terms. Their genuine experience — perhaps more relevant than the person who optimised — is described in their own words rather than the job description’s language. The ATS ranks them lower. They don’t make the shortlist.

Meanwhile, a weaker candidate who followed a three-step AI sequence gets through, because their document perfectly mirrors what the screener was looking for.

This is not a hypothetical. It’s the structural consequence of a screening system that evaluates documents rather than people. AI didn’t create that flaw — it’s always been there, which is why resume writing as an industry has existed for decades. AI has simply made it available to everyone, at scale, for free. The gap between the candidate who knew how to present themselves and the one who didn’t just got much narrower and much cheaper to cross.

The Blur Between Coaching and Fabrication

There’s one more dynamic worth naming. The line between “surfacing genuine experience better” and “embellishing experience that isn’t quite there” is narrower than it looks, and AI makes it easy to cross without fully noticing.

Prompt 3 says “while telling the truth.” But what happens when an AI asks a clarifying question and the candidate gives an answer that stretches slightly, and the AI writes it up in a way that stretches further? What happens when “I contributed to a project that improved team efficiency” becomes “drove a cross-functional process improvement initiative that delivered measurable efficiency gains”? The candidate didn’t lie exactly — but the resume no longer represents what actually happened.

This is the spectrum we described in Part 1: from polished, to embellished, to fabricated. AI-assisted career coaching sits at the legitimate end of that spectrum. The problem is that it’s the same set of tools, the same workflows, and often the same platforms as the ones producing the fabricated end. And the output, on paper, looks identical to a hiring manager working from a document.

What This Means for Employers

The arms race framing holds, but it’s worth being clear about what employers are actually racing against.

It’s not primarily fraudsters. It’s a mainstream shift in how candidates interact with hiring processes — one that’s being openly taught, widely adopted, and largely legitimate in intent. Treating it as a bad-faith attack misses the point and will lead to unfair outcomes for candidates who are doing nothing wrong.

What it does mean is that keyword matching — whether done by a human reading a resume or an ATS doing it automatically — is a weaker signal than it used to be. A resume that perfectly mirrors a job description now tells you less about fit than it once did, because the mirroring may be the product of a prompting sequence rather than genuine alignment.

The response isn’t suspicion. It’s a different kind of evidence.

A candidate’s DISC behavioural profile doesn’t shift based on how they prompted an AI. Their structured interview responses don’t mirror a job description they read beforehand. The consistency of their answers under follow-up questions — probing for specifics, dates, colleagues, outcomes — doesn’t change because they optimised their document. These are the layers of a hiring process that can’t be gamed by knowing the right TikTok to follow.

The GrowMyTeam Approach

GrowMyTeam was built around the recognition that a resume was always one data point — useful, but limited, and the easiest part of the process to manipulate. That was true before AI. It’s more true now.

Every application that comes through the platform is scanned for AI-generated content, embellishment, and hidden manipulation attempts before any assessment begins. But the more significant shift is what happens next: validated DISC behavioural profiling builds an independent picture of the candidate that exists entirely outside what they chose to write about themselves. Structured interview data adds a third layer. By the time a hiring manager makes a call, they’re working from several independent sources — not one document, however cleverly optimised.

The candidate who followed this kind of three-prompt sequence may have a polished, well-targeted CV. They may also be an excellent hire. The process should find out which — not assume either.

That’s what fair assessment looks like when the arms race is running: not a harder filter on the document, but a process that was never built to depend on the document alone.

GrowMyTeam.ai scans every application for AI-generated content, embellishment, and hidden manipulation attempts — then builds the real picture from validated behavioural profiling and structured interviews, so hiring managers are working from substance, not spin. Book a demo or try our free profiling tool.