I’m finding that most CVs I receive look the same now.
Same clean layout. Same confident summary. Same list of “spearheaded”, “aligned” and “orchestrated”. Tools that rewrite your CV in thirty seconds are everywhere, and a lot of people are using them. Fair enough. The CVs look professional, are well written, but it turns out to be a case of style over substance.
If you hire in IT, digital, accounting, finance, risk, or corporate, you already know the feeling. Ten CVs land. All of them read well. Three of them fall apart the moment you ask what the person actually did.
This is not a guide to beating an applicant tracking system (ATS). This is about the content: how to tell a real CV from an AI-smoothed one, and how to write yours so it still sounds like you.
What AI CVs usually Looks Like
The same phrases keep showing up. The summary could belong to anyone in the profession. “Results-driven professional with a passion for continuous improvement” tells you nothing about the last job.
Every bullet starts with a strong verb and ends with no number, no system, and no scope. “Led cross-functional initiatives to deliver outcomes” is the giveaway.
The language is oddly even. No rough edges. No specific product names. No messy South African context. It reads like it was written for a global template, not for a bank in Johannesburg or an engineering company in Boksburg.
Keywords from the ad appear once each, neatly, like they were sprinkled on top.
None of that proves someone used AI. Good writers can generalise as well, but you can spot the signs.
If you are the hiring manager
Do not ban AI. You will not win that fight, and a candidate who tidies their grammar is not the enemy.
Change what you reward.
Ask for proof in the first screening call. “What was the budget, the team size, the system, and your bit?” If the CV cannot answer that, the interview will not either. Put two or three knockout questions on the application that only someone who did the work can answer confidently.
In the interview, pick one bullet and go deep. “You said you delivered the finance transformation. Walk me through week one.” People who did the work get specific. People who polished a story get stuck on the adjectives.
Watch for identical phrasing across candidates for the same role. If three CVs use the same unusual sentence, something copied the brief back at you.
And write a better brief yourself. A vague ad produces vague CVs. If your must-haves are soft, AI will fill them with soft language. Be specific about the work, and you will get specific people.
If you are the candidate
Use the tools if they help you edit. Then put the human back in.
Lead with one line only you could write. Scope, scale, outcome. Keep the strong verbs. Attach them to real things. System names. Rand amounts. Headcount. Deadlines. Regulators. Customers. If you cannot add a number, add a constraint: what broke, what you chose, what you would not do again.
Mirror the ad with language you actually earned. Do not paste every keyword. If the ad asks for Dynamics and you have Dynamics, say where and for how long. If you do not, do not invent a sentence that sounds like you do.
Read the CV out loud. If it does not sound like something you would say in a room, cut it. That is the fastest AI-polish detector I know.
For the screening-tool side of this, keep formatting clean and put the important evidence early.
Hiring managers: can you explain, from the CV alone, what this person owned in their last role? If not, it is not ready for a shortlist.
Candidates: can a stranger tell, in thirty seconds, what you are good at and where you did it? If not, the polish is hiding you.
AI did not invent empty CVs. It just made them faster to produce. In a market where inboxes are already busy, the people who win are still the ones who can point to the work.
If you need help constructing your CV, or help working out how to filter out the real applications for the AI-generated ones, get in contact.
