The CV gets read in 11 seconds. The machine decides in 3.
Neither of those things are entirely true. But both of them affect you.
of people have used or plan to use AI in their job search in 2026. LinkedIn, Jan 2026, n=global panel.
of hiring managers believe they can tell when someone used AI on their application. Insight Global, 2025. Key word: believe.
have encountered AI-generated applications. Detection rate climbed from 53% to 77% across two years. CoverSentry, compiling 20+ surveys, 2026.
of employers reject AI CVs that lack personalisation. Resume Now survey, n=925 HR professionals, March 2025.
"The employer uses AI to screen your CV. You used AI to write it. And 67% of companies admit their AI screening tools could introduce bias. We are in a very expensive arms race with absolutely everyone losing money on the entry fee."
Source: Resume Builder survey of 948 US business leaders, Oct 2024. SHRM 2025/2026 surveys. CoverSentry 2026 data compilation.
Employers using AI to screen CVs while hating that candidates use AI to write them isn't contradiction. It's a signal problem. Here's what they're actually worried about.
A CV exists to answer one question: can this person do this job? AI produces text that reads like someone who can do every job. It's not lying, exactly. But it's noise where there should be signal.
78% of hiring managers say personalised details signal genuine interest and fit. An AI CV is a personalised impression of a generic candidate. The bit they're looking for you isn't there. Source: Resume Now, 2025.
64% of recruiters already saw more look-alike applications in 2024–25 thanks to GenAI. ResumeBuilder/ZipRecruiter, 2025. The same recruiter who used to get 100 CVs now gets 500. All of them start with "results-driven professional."
First example: AI-generated. Second example: human-written, specific, verifiable. One of these gets callbacks.
A National Bureau of Economic Research randomised controlled trial of 480,948 job seekers found AI writing assistance increased hiring by 7.8%. The key word is assistance AI used to sharpen human-written content, not replace it. NBER Working Paper 30886, 2023. Still the best-controlled study on the topic.
33.5% of hiring managers say they can spot an AI-generated CV in under 20 seconds. Not because they have superpowers. Because AI makes the same mistakes, every time. TopResume survey of 600 hiring managers, May 2025.
ChatGPT loves this phrase. So does Claude. So does Gemini. If your CV starts here, every recruiter's stomach sinks a little.
AI produces text that's grammatically perfect but lacks a human voice. No anecdotes. No texture. No sense of a person behind the words.
"Led successful projects" what projects? What counts as success? AI creates text that sounds nice but says nothing. A real candidate says: "Delivered 12-market rollout, £2.4m under budget, Q2 2025."
AI inaccurately attributes skill levels. "Expert in crisis management" for someone two years into their career. Recruiters know how skills develop. The mismatch is a glaring flag.
Formal in one paragraph ("optimised operational workflows"), casual in the next ("really excited to contribute"). Different AI prompts, different results. One CV, multiple voices. Human CVs don't do this.
The cover letter doesn't mention the company name. Or the role. Or the hiring manager. Because you made one prompt and sent it 40 times. They know. 53% of hiring managers are frustrated by impersonal outreach. Resume Now, 2025.
Humans make choices. They emphasise some things and underplay others. AI tries to include everything because it doesn't know what matters. A CV that covers every possible skill is a CV written by something with no skin in the game.
You wrote "expert in Salesforce architecture." They ask you one detailed question. You pause for three seconds. That's the tell. The CV lied. You didn't mean to. But here you are.
The world's first comprehensive AI law entered into force August 2024. For recruitment specifically, the critical deadline is August 2, 2026. As in: this month. European Commission, Regulation (EU) 2024/1689.
The regulation is law. The clocks started.
Emotion recognition in hiring: illegal. Biometric categorisation by protected characteristics: illegal. If you were using these, you've already broken the law. Source: Article 5, EU AI Act.
CV screeners, candidate ranking, interview AI, performance prediction all classified as high-risk. Mandatory: bias testing, human oversight, audit trails, transparency with candidates. Sources: Articles 9–15, EU AI Act; Annex III, Section 4.
Up to €35 million or 7% of global annual turnover. Whichever is higher. Source: rexx-systems.com citing EU AI Act penalty framework.
Source: Articles 9–15, EU AI Act (Regulation EU 2024/1689). Hiretruffle.com compliance guide, April 2026. Boundlesshq.com, March 2026.
This applies to any company using AI to hire people in the EU, even if the company is based in the US, UK, or anywhere else. You hire one person in Germany using an AI CV screener? Welcome to compliance obligations.
Source: Hiretruffle.com EU AI Act guide, April 2026 confirmed by European Commission guidance.
The EU AI Act doesn't apply in the UK post-Brexit. The UK is pursuing a "pro-innovation" framework without a single binding AI law yet. Regulators are developing sector-specific guidance. Watch this space. Source: Boundlesshq.com, March 2026.
"Employers are about to be legally required to audit the AI they use to read your CV for bias, for transparency, for human oversight. Meanwhile they're rejecting you for using AI to write it. If you're a candidate, this is your moment to say: 'tell me about your AI compliance programme.' Watch them squirm."
That last bit is a joke. Mostly.
ATS software is used by 97.8% of Fortune 500 companies. Jobscan ATS Usage Report, July 2025. Skills-first hiring is now the standard, with 43% of businesses making it their top priority. The Interview Guys, May 2026. Here's what actually works.
The viral claim that "75% of CVs are auto-rejected before a human sees them" traces back to a sales pitch from a startup that went out of business in 2013. There is no primary source. It was invented to sell resume services. The actual risk is real but different: badly formatted CVs parse wrong. JobCannon.io, 2026, citing research trail.
Complex designs, tables, headers and footers, two-column layouts and decorative icons all cause ATS parsers to misread or lose your content. You're not being rejected. You're just invisible. Use clean, single-column layouts for any application portal. Keep two versions one ATS-safe, one beautiful for direct email.
With 43% of employers now leading with skills over credentials, your degree from 2009 is doing less heavy lifting than it used to. What does the work: certified skills, measurable outcomes, specific tools and technologies you've actually used, portfolio links that work. Source: The Interview Guys, citing ApplyGlide and LinkedIn data, 2026.
The format has evolved. Skills-first, ATS-compatible, metrics-forward, human-proof. Here is what works in 2026 for machines and for people. Sources: Schulmeister Consulting, March 2026. PE Global, Dec 2025. The Interview Guys, May 2026.
"In 2026, AI literacy has become a baseline expectation. Include specific AI tools you use and how you leverage them for efficiency." The Interview Guys, May 2026. Don't list "ChatGPT." List what you used it for and what it produced.
Not because of an arbitrary rule. Because attention is scarce and your job is to make it easy to say yes to you.
"Seeking an exciting opportunity to leverage my skills in a challenging role where I can contribute to a dynamic team." This is what AI writes. This is also what 80% of people write. Don't. Tell them what you specifically want and why this role fits. If you can't write it without sounding like an AI, that's your cue to think harder about why you actually want this job.
Only about 37% of employers now see traditional credentials as reliable talent signals. Over 40% are actively ditching resume-first hiring. Willo Hiring Trends 2026. The reckoning is coming. But it's not here yet.
Tech, creative, engineering portfolio-first hiring is already standard in many organisations. Some companies have replaced CVs with skills tests entirely. If you're applying to a startup and you haven't shipped anything, a CV won't save you.
Finance, law, healthcare, professional services, the public sector the CV is alive and the interview process is still largely sequential. The shift is slower here. A bad CV will still cost you. But "bad" now means generic and unverifiable, not two columns.
The CV needs to become a proof document, not a biography. Your job history is context. Your results are the product. Add a portfolio link, a GitHub, a case study, anything that shows rather than tells. The CV plus something beats a CV alone.
Employers anticipate that 39% of key skills required in the job market will change by 2030. Your 2018 CV with "Microsoft Excel" under skills is doing about as much work as a fax machine. Update it. Then update it again. Source: World Economic Forum Future of Jobs 2025, cited in recruitment.com analysis, March 2026.
93% of recruiters plan to increase AI use in 2026. LinkedIn is often the first search, not the job board. If your profile contradicts or undersells your CV, that's a problem before they've even read the document. Source: LinkedIn Talent Solutions, January 2026.
80% of employers are uneasy seeing AI-generated cover letters. They're optional more often now. When you do write one: three paragraphs. Why this role. Why now. What specific thing you bring that the job description is clearly looking for. CrownStaffing, January 2026.
AI-conducted interviews tripled in two years from 10% to 34%. Two-thirds of recruiters plan to expand AI pre-screening in 2026. Your CV is now just the warm-up. CoverSentry, 2026, citing ResumeBuilder data.
66% of Americans say they wouldn't apply to an employer that uses AI for final hiring decisions. Pew Research, cited in SQ Magazine, 2026. You're allowed to ask how decisions are made. Under the EU AI Act, in Europe, they're legally required to tell you.
No waffle. No hedging. The ten things that actually change outcomes.
Then use AI to sharpen it. Not the other way around. The experiences are yours. The voice should be yours. AI is the editor, not the author.
If you can't attach a metric to something, describe the scale, the impact, or the before-and-after. "Led team" → "Led eight-person team delivering £1.4m project three weeks early."
62% of AI CVs get rejected for lack of personalisation. One prompt, 50 sends = 50 rejections. Spend 20 minutes per role swapping in their language, referencing their context.
Experience. Education. Skills. Don't get creative with section titles. ATS parsers are literal. "Where I've Made My Mark" is not a section an algorithm recognises.
Dead LinkedIn link = first impression gone. Outdated GitHub = worse than no GitHub. Broken portfolio = interview question you don't want. Check everything before you send anything.
AI will helpfully add "crisis management" to your skills section. If you're two years into your career, a five-minute interview question will expose that claim. Only list what you could prove in a room.
This is the single most effective test for AI content. If it sounds like a press release, it will read like one. If it sounds like someone explaining their job to a sensible person, you're good.
A case study. A portfolio piece. A published article. A GitHub repo. A certification. Anything that proves you do what your CV claims. The CV plus evidence beats a CV alone, every time.
You will be asked about everything on it. "Tell me more about that project." If you can't, because AI wrote it and you submitted it without really reading it, the interview is over in the first five minutes.
The worst time to write your CV is when you desperately need a job. The best time is three months before you're looking, when you can be thoughtful, accurate and not panicky. Update it quarterly. Even one bullet point per role, per quarter.
