You Can’t Spot a Lie on a CV by Reading It Harder. Here Is What Actually Works.

Almost Every CV Rounds Up

Start by lowering the temperature. Embellishment is not a fringe behaviour you occasionally run into; it is the water the whole exercise swims in. Surveys of French candidates put the share who admit to stretching or dressing up their CV at more than one in two. Most of it is soft: a generous job title, a solo project reframed as a team win, a tool used twice listed as a core skill. People are not plotting against you. They are writing the document the format asks for, which is the strongest honest case they can make for themselves.

It matters anyway, because some of that rounding lands on things the job actually depends on. In a Robert Half survey of 300 company directors in France, 47% said they had already ruled out a candidate after finding false or exaggerated information in the application. The same study found the inflation concentrated in a few predictable places: prior experience, diplomas and qualifications, technical skills, and languages. None of that means candidates are dishonest as a rule. It means the CV is an argument, and arguments round up.

Worth separating from all of this is the rarer, sharper problem of a candidate who is not who they say they are at all, which we cover in our piece on identity fraud and deepfakes in remote hiring. That is a different threat with a different fix. Everyday embellishment is not fraud; it is the ordinary gap between how good a CV reads and how much of it is load-bearing, and it is the gap most of your hiring time gets spent on.

Why Reading Harder Doesn’t Catch It

Here is the uncomfortable part: the better a CV is written, the more convincing it is, and being convincing has almost nothing to do with being true. When you read a stack at speed, you reward confident phrasing, clean formatting, and the keywords you were hoping to see. Those are exactly the things an embellished CV optimises for. The candidate who undersells, who writes “helped with” when they led, reads as weaker than the one who writes “drove” when they watched. You are not misjudging people; you are grading a document that was built to be graded well.

This gets worse under volume, and it is a close cousin of the problem we described in the real cost of screening CVs by hand. The faster you go, the more you lean on surface signals, and surface is precisely where embellishment lives. Reading with more suspicion does not fix it either: it just adds anxiety to a task that still has no way to confirm anything. The failure is not a lapse of judgement. It is structural. A CV is a claim, and a claim cannot be checked against itself, no matter how carefully you stare at it.

The Claims Worth Verifying

The answer is not blanket suspicion; that is slow and it poisons the candidate experience. It is targeting. Most of what gets rounded up does not matter, and a few things do. The trick is to know which claims are load-bearing for this specific role, and to spend your verification effort only there. Roughly, the CV breaks into claims you can safely judge later and claims worth confirming before you commit.

Read down that last column and the pattern is clear. A rounded-up title is the genre and rarely worth a fight; an invented diploma or a fabricated result sits on something the role actually depends on. Deciding which is which is not a character test of the candidate. It is a question about the job: what does this person genuinely have to be able to do, and which of those things is the CV asking me to take on faith?

Verify Against Evidence, Not the CV

Once you stop trying to authenticate the page and start gathering independent evidence, the whole problem softens. The CV stops being the verdict and becomes a hypothesis, and every stage of your process is a chance to test it against something that is harder to dress up.

• Ask for specifics, not summaries. A claim survives when someone can walk you through what they actually did; it dissolves under “what was your part exactly, and how did you know it worked?”. This is the whole case for structured, behavioural interviews: the same evidence-seeking questions for everyone.
• Test the one skill the job runs on. A small, realistic piece of work is the one exercise a padded CV cannot pass. It does not need to be long; it needs to be the actual thing the role does.
• Point a reference at a specific doubt. A general “was she good?” confirms what you hoped; a narrow question aimed at the one claim you are unsure of tells you something, as we argue in how to make a reference check say something.

The thread joining these is that no single step decides anything. You score each stage against the same criteria, fixed before you met anyone, so a fluent CV and a plain one are read on the same evidence rather than on how well they were written. That is a large part of why we built Kynto to score candidates against the criteria you define across the whole process, not on how the CV reads, and to keep a person reading the reasoning behind every score. The point is not to distrust candidates. It is to stop asking a sales document to be a source of truth, and to check the few claims that matter against work you can actually see.

Key Takeaways

• Embellishment is the norm, not the exception. More than one in two candidates admit to it, and 47% of French directors have ruled someone out over false or exaggerated information. A CV rounds up because that is what the format is for.
• You cannot catch it by reading harder. Scanning at speed rewards confident writing, which is exactly what an inflated CV optimises for. The failure is structural, not a lapse of judgement: a claim cannot be checked against itself.
• Target the few load-bearing claims and check them against evidence: specific behavioural questions, a short work sample, a reference aimed at a real doubt, all scored on the same criteria. The CV becomes a hypothesis, not the verdict.

FAQ

How do I know if a candidate is lying on their CV?

Mostly, you do not, from the CV alone. Treat every important claim as a hypothesis and check it against evidence: a specific behavioural question, a short work sample, a targeted reference. The claims worth the two minutes are the load-bearing ones, a required diploma, a skill the job runs on, a headline result, not a generously worded job title.

Should I reject a candidate who exaggerated on their CV?

It depends on what, and how much. A rounded-up title or generous framing is the genre, not a red flag. An invented diploma or a fabricated role is a different matter, both because the claim was material to the job and because it tells you how someone handles the truth under pressure. Judge the substance, not the polish.

Isn’t it faster to trust the CV and interview well?

A good interview is exactly how you verify, but only if it is built to test claims rather than to enjoy a conversation. A free-flowing chat rewards the same fluency the CV does. Specific, evidence-seeking questions asked of every candidate are what turn an interview into a check instead of a pleasant hour.