How to Detect an AI-Written CV (And What It Means for Hiring in 2026)
In 2026, a significant proportion of every recruiter's inbox is AI-generated. Some estimates put it above 40% for technology and professional services roles. The question most companies are asking is: "How do we detect it?" The better question is: "Why does it matter - and what should we do instead?"
Why Candidates Are Using AI to Write Their CVs
Before judging the practice, understand the context: candidates are applying to roles through an ATS that screens for keywords, not quality. The fastest way to match those keywords is to feed the job description to an AI and ask it to tailor your resume. The system is incentivising exactly the behaviour companies are complaining about.
Most AI-generated CVs aren't fraudulent - they represent a real person's genuine experience, described in language optimised for the screening software. The content is true. The presentation is machine-generated.
7 Signals of an AI-Written CV
- Perfectly mirrored job description language. Every key phrase from the JD appears verbatim in the resume. Human candidates paraphrase. AI tools copy and embed.
- Uniformly perfect prose. Every bullet is the same length, the same structure, the same voice. Real career histories have variation - different eras, different writing styles, different levels of detail.
- Generic superlatives without specifics. "Results-driven professional with a proven track record of delivering exceptional outcomes" - with no actual outcomes mentioned. AI generates confident language around vague claims.
- Suspiciously comprehensive skill coverage. The candidate happens to have deep experience in every single requirement listed in the job description. All of them. At exactly the right depth.
- No personality or distinctive voice. Human resumes have quirks - an unusual project, an unexpected detail, a career decision that needs a sentence to explain. AI-generated ones are frictionlessly generic.
- Inconsistent specificity. Some bullet points are highly specific with real numbers ("increased ARR by 34% in Q2 2025"), while others are vague generalities. A mix of real details and AI-generated filler.
- Cover letter that reads like a press release. Formal, well-structured, hits all the right notes, and could have been written for any company in the same sector.
Important nuance: These signals indicate AI assistance, not dishonesty. A candidate who used AI to improve the presentation of their genuine experience is not being fraudulent - they're adapting to the incentives your screening process created. The real question isn't "did they use AI" but "can they actually do the job?"
Why Banning AI CVs Misses the Point
Some companies have introduced policies banning AI-assisted applications. This is largely unenforceable and misses the real issue: the problem isn't that candidates used AI. The problem is that your screening process can't distinguish between strong and weak candidates regardless of how their CVs were written.
An AI-generated CV from a brilliant candidate still represents a brilliant candidate. A human-written CV from a poor fit is still a poor fit. The tool used to write the document is less important than what the document represents - and far less important than what the person can actually do.
What Smart Companies Are Doing Instead
Rather than trying to detect and ban AI applications, the most effective companies in 2026 are redesigning their screening process so that AI-generated text becomes irrelevant:
- Asynchronous video screening: A 2-minute video answering specific questions about the role reveals communication ability, genuine interest, and personality that no AI can fabricate
- Skills assessments: A relevant take-home or timed task shows what the candidate can actually do, independent of how their resume was written
- Portfolio and project evidence: Real demonstrated work - GitHub, published writing, design portfolios, case studies - can't be faked by AI the same way text can
- Referral-based sourcing: Candidates who come via warm introductions from trusted sources arrive with a human signal of quality that AI-generated applications can't provide
Starity: Making AI CVs Irrelevant
On Starity, candidates don't just submit resumes - they build career profiles from real project outcomes, demonstrated work, and verified experience across 11 dimensions. An AI-generated description of skills is replaced by actual evidence of those skills in action.
Recruiters on Starity know exactly who they're meeting before the first call - not because they read a perfectly optimised CV, but because they've seen real work, real outcomes, and a real career trajectory. Proof, not just claims.
The AI CV problem disappears when your hiring process is designed around evidence, not text.
Hire people for what they've actually done - not what their CV says
Starity replaces keyword screening with career intelligence. Real outcomes. Demonstrated work. Actual fit. Join the waitlist.
Join the Waitlist