Recent research by workplace engagement platform Kahoot! revealed that 65% of graduates are now using tools like ChatGPT to write cover letters and job applications. Over half (51%) also admitted to overstating their skills on their CV or LinkedIn profile, with work experience being the most embellished area where 53% acknowledged exaggeration.
Graduates also reported inflating language skills (39%), education or qualifications (37%) and job titles (36%).
Therefore, we all now - whether recruiter, HR team member or hiring manager - need to look beyond the polished CV and consider the broader ecosystem jobseekers operate in, one where AI-assisted applications are becoming the norm.
A recent poll we conducted on Reed’s LinkedIn profile, revealed that only 37% of organisations have adapted their hiring process to account for AI-generated and over-exaggerated CVs and cover letters. As recruitment experts, we’re seeing a shift in how candidates approach the job market. Gone are the days when a CV was purely a reflection of someone's late-night typing efforts. Today, AI tools like ChatGPT have made it incredibly easy for applicants to generate polished, professional-looking documents in seconds.
While this technology can level the playing field for those who struggle with writing, it poses a distinct challenge on the hiring side. Spotting AI usage isn't about catching people out or disqualifying them immediately. It is about understanding who you are really evaluating. By recognising the signs of machine-generated text, you can tailor your interview process to dig deeper into the applicant’s true capabilities.
In this article, a handful of our recruitment experts highlight some tell-tale signs a CV or covering letter might have had a digital ghostwriter.
1. The language is flawless but generic
One of the first things you might notice about AI-generated text is its perfection. AI models are trained on vast datasets of professional writing, meaning they rarely make grammatical errors. Alison Whitmore-Ryan, Area Manager for Reed’s education division in Maidstone, says: “While a typo-free application is usually a good sign, AI takes this to an extreme. I have worked in recruitment for years and never have I seen so many CVs without grammatical mistakes.”
When talking further on this topic, Alison adds: “On paper, AI makes people seem extremely creative, it almost gives them a poetic licence. But when you talk to and meet the candidates, that’s when you know for certain that their CV was either written completely by, or enhanced by, AI.”
Matthew Stirling, Recruitment Consultant specialising in accountancy recruitment, in Oxford, said: “You should look out for text that feels sterile. It reads like a textbook definition of the job role rather than a personal account of experience - I have seen more and more of this on CVs. The sentences are often perfectly balanced and use sophisticated vocabulary, but they lack the idiosyncrasies of human speech.
“For example, a human might write, "I managed a team of five and we decreased profit and loss by 20%." AI might embellish this to, "Orchestrated a dynamic team of five professionals, leveraging synergistic strategies to drive a 20% decrease in profit and loss." It is technically correct, but it lacks the authentic voice of a real person. If the application reads like it was written by a corporate robot, it probably was.”
2. A disconnect between CV and cover letter
Consistency is a hallmark of authentic writing. When a person writes their own application materials, their unique ‘voice’, their choice of words, sentence structure, and tone, usually permeates both documents. When AI gets involved, you often see a jarring disconnect.
Perhaps the covering letter is a masterpiece of eloquent persuasion, filled with complex sentence structures and emotive language, while the CV is bullet-pointed, dry, and uses a completely different vocabulary set. This often happens if an applicant uses AI to generate the letter but updates an old CV manually (or vice versa).
As an employer, pay attention to these shifts in style. Does the person describing their passion in the letter sound like the same person listing their duties in the CV? If the two documents feel like they were written by two different people, one of those ‘people’ might be AI.
3. American spellings in a British content
Most leading AI text generators are trained predominantly on US English data. Unless the user specifically prompts the AI to write in British English (and many people often forget to do this, or even change this), the output will default to American spellings and terminology.
“If you are recruiting for a role in Oxford or anywhere in the UK for that matter, seeing "optimized", "color", "center", or "program" (instead of "optimised," "colour," "centre," or "programme") can be a subtle giveaway,” says Matthew. “You might also spot Americanisms in phrasing, such as "in the fall" instead of "in autumn," or references to "Grade Point Average (GPA)" rather than degree classifications.”
While some may simply prefer American spellings or have worked internationally, a CV riddled with US terminology from an individual who has lived and worked exclusively in the UK is a strong indicator of AI assistance.
4. Vague achievements lacking context
AI is fantastic at sounding impressive without actually saying anything concrete. It knows that a "results-oriented professional" sounds good, but it doesn't know the specific details of someone’s career unless it's been fed them very carefully.
Consequently, AI-generated CVs are often stuffed with vague, high-level claims that lack specific metrics or context. You might see bullet points like:
"Improved departmental efficiency through strategic implementation of new protocols," or,
"Enhanced customer satisfaction levels by fostering a culture of excellence."
These sound great, but what do they mean? A human writer is more likely to include the messy, specific details: "Cut processing time by two days by introducing Trello for task tracking," or "Raised Trustpilot score from 3.5 to 4.8." If the achievements float in a sea of abstraction without anchoring facts, you are likely reading AI output.
5. Repetitive sentence structures and the em dash
As humans we naturally vary our sentences to keep things interesting (I sincerely hope I have managed to keep you engaged up until this point). We mix short, punchy sentences with longer, descriptive ones. AI, on the other hand, often falls into repetitive patterns.
Gillian Dolan, Regional Director at Reed advises: “When looking at CVs, scan the document for structural repetition. Does every bullet point start with a verb? "Managed...", "Created...", "Led...", "Developed...". While this is standard advice for CVs, AI tends to overdo it, often using the same ‘power verbs’ repeatedly.
“Also, look for paragraphs that all start with a similar transitional phrase, such as "Furthermore," "Additionally," or "However." This formulaic approach to structure is a key signature of algorithmic writing.”
But, on top of this, there is one grammatical ‘tell’ for identifying text that has been generated by AI, that’s the em dash. Gillian adds: “Often in candidate comms and especially CVs you will see the extra-large dash in the text. AI appears to use them frequently as a way to break up text, like they are a shortcut for creating complex sentence structures, and you can spot them a mile off.”
6. Over-optimisation for ATS keywords
Some jobseekers may know that employers use applicant tracking systems (ATS) to filter CVs. AI tools are notoriously good at shoehorning keywords into CVs to beat these systems.
If you are reading a CV and it feels like a word search of your own job description, be suspicious. You might find sentences that feel forced or clunky because they are trying to cram in terms like "stakeholder management," "agile methodology," and "cross-functional collaboration" all at once.
Those writing authentically will weave these skills into their narrative naturally. AI will often list them or force them into sentences where they don't quite fit, prioritising the algorithm's approval over readability.
7. The hallucination of skills and experiences
This is the most dangerous sign for employers. AI models can sometimes, what we would call ‘hallucinate’ - they make things up to fill gaps or satisfy a prompt, especially where they don’t have enough information given to them from the off. If an individual asks AI to "write a CV for a senior marketing manager role,” without additional insight, the platform might invent plausible-sounding projects or skills that the person doesn't actually possess.
Kimberly Hunsdon, Divisional Manager and engineering recruitment specialist based in Southend, said: “People sending over their CVs certainly appear to have more qualifications than ever before. On top of this, I sometimes spot skills that are slightly off or outdated, or references to software that doesn't quite exist in the way described. For example, listing proficiency in an engineering programming language that isn't relevant to the role, or claiming mastery of a tool that was only released last month with years of experience.
“Always verify the details. If a project description sounds generic or slightly off for your industry, it warrants a closer look during the interview stage.”
How employers should respond
So, you have spotted a CV that screams ChatGPT. What now?
Using AI isn't necessarily a bad thing. It shows resourcefulness and an ability to use modern tools. In fact, for some roles, being able to prompt an AI tool effectively is a desirable skill.
However, you need to verify the person who is applying for your job, has the actual ability to do the job you want them to do.
Test for substance in the interview or a screening call: If the CV is vague, drill down into the specifics. Ask, "You mentioned you 'spearheaded a strategic initiative', can you walk me through exactly what you did, day by day, to make that happen?"
Use practical assessments: Move beyond the CV. Give interviewees a short, real-world task relevant to the job. If they used AI to write the application, they can't use it to fake their way through a live coding test or delivering a presentation.
Check for communication skills: If the role requires writing, ask for a writing sample to be produced during the interview process, or simply gauge their verbal communication against the eloquence of their cover letter.
Ultimately, your goal is to find the best people for your company. AI is a fantastic tool in the modern workplace but learning how to spot the signs can help you look past the polished surface to see the real potential underneath.
Our recruiters are well-versed in spotting AI-crafted content. So, if you’re looking to hire well, get in touch with one of our local specialists today.
