Somewhere right now, a registered manager is scrolling through 200 applications for a single care worker role. Most of those CVs were written, rewritten, or polished by an AI. The manager knows this. The applicants know this. And if the provider is large enough to run an applicant tracking system, the software reading those CVs is also an AI. We have built a closed loop where machines write documents for other machines to read, and somewhere in the middle a human being is trying to work out who to trust with their residents.

It's a bit absurd when you say it out loud.

What has AI actually changed about the CV?

Five years ago, writing a CV was a genuine skill. You had to think about what you'd done, figure out which bits mattered, and find the words to make it sound like you rather than a LinkedIn horoscope. It was tedious, yes, but the tedium was the point. The act of writing forced people to reckon with their own experience.

Generative AI collapsed that process in about eighteen months. Paste in a job description, paste in an old CV, and out comes a tailored, keyword-stuffed, bullet-pointed version that sails through most filters. The barrier to applying for a job fell from "an afternoon" to "ninety seconds." Predictably, application volumes went vertical. Some care providers now report two to three times the number of applications per role compared to 2022 — and the quality signal inside those applications has gone down, not up.

Recruiters, already overstretched, have responded by leaning harder on their own AI filters. And so the arms race began.

The uncomfortable truth

Here's the part the careers-advice industry doesn't want to admit: the CV was never a great document in the first place. It was a compressed, standardised, mildly dishonest summary of a person, optimised for a world where hiring managers had time to read it carefully. In care hiring, that world never really existed. Registered managers have always been hiring between rota gaps, between CQC paperwork, between a 6am call-out and a 9pm medication round.

AI didn't kill the CV. It just exposed how fragile the whole ritual always was.

Let's be honest about the candidate side too. Using AI to polish a CV isn't cheating. It's the same instinct that led people to pay CV consultants £200 a pop in 2015. The tool got cheaper and better; the impulse is identical. Telling a care worker — often juggling a second job, a family, and a commute — to stop using AI is like telling them to stop using spellcheck. It's not going to happen, and moralising about it helps no one.

So what is the actual problem?

The problem isn't that AI-written CVs are somehow inauthentic. The problem is that when every CV is buffed to the same mirror shine, the CV stops carrying any signal at all. If every applicant "drove cross-functional alignment to deliver 30% efficiency gains," none of them did. The document has become noise.

And the costs of that noise land hardest in sectors like care, where hiring the wrong person isn't an inconvenience — it's a safeguarding risk. CQC doesn't care how good someone's CV looked. CQC cares whether you had a structured, fair, evidenced process for deciding they were right for the role.

What care providers should actually do

Stop pretending the CV pile is a meritocracy. If you're using keyword matching to screen out most applicants, you're not hiring the best — you're hiring the people whose CVs were best at impersonating your job description. That's a compliance problem as much as a quality problem.

Three practical shifts:

Move the first real conversation earlier. Structured phone screening — the same questions, scored the same way, for every applicant — is fairer and more revealing than any CV. Lily runs this 24/7, so applicants get a call within hours of applying instead of waiting a week.
Treat the CV as a starting point, not a filter. Use it to prompt questions, not to eliminate candidates. A patchy CV often hides a strong care worker who changed careers, took time out for family, or simply doesn't write about themselves well.
Build a proper audit trail. Every screening decision should be documented, consistent, and defensible if CQC asks. This is the part most providers get wrong — and the part most at risk when AI is doing the filtering invisibly.

What candidates should actually do

Use AI. Of course use AI. But use it the way a decent writer uses an editor, not the way a lazy student uses a homework mill. Let it catch your typos, tighten your phrasing, suggest verbs you'd forgotten. Don't let it invent experience you don't have, don't let it launder your voice into corporate mulch, and for the love of everything, read the output before you send it — "Would you like me to redraft this in a more professional tone?" is still landing in hiring managers' inboxes more often than you'd think.

More importantly: invest the time you've saved somewhere that still matters. When Lily calls you, answer honestly. A specific, concrete story about a resident you supported will always beat a bullet point about "delivering person-centred care outcomes."

The numbers behind this

At Sonderwell, moving from CV-led screening to structured AI phone screening produced 127% more hires per month and 616 admin hours saved — over a 12-month period, with 2,464 applicants processed (Sonderwell case study, 2024). Time-to-screen dropped from the UK care sector average of 9 days to 0.83 days. The interesting part isn't the speed. It's that 91% of candidates rated the experience positively and 92% of applicants got a response. Nobody was ghosted. Nobody was filtered out by a keyword.

That's what replaces the CV as the first line of judgement: a fair, structured, recorded conversation. Not a document.

The real question

AI hasn't ruined CVs. It has revealed that the industry has been performing a ritual that stopped making sense some time ago. The interesting question isn't "how do we detect AI-written applications" — that ship has sailed, and the detectors don't work anyway. The interesting question is what replaces the CV when it finally collapses under its own weight.

In care, we think the answer is already here. A structured phone call, the same for every applicant, scored consistently, with a proper audit trail behind it. CQC-ready. Candidate-friendly. Faster than any ATS.

Something will replace the CV. Something has to. And honestly? Good.

FAQs

Is it fair to use AI to screen care worker applications?

It can be more fair than the alternative, if it's done well. Structured AI phone screening asks every applicant the same questions and scores them the same way. That removes a lot of the unconscious bias that creeps in when humans skim CVs under time pressure. The key is transparency, consistency, and a human making the final decision.

Does CQC accept AI-led screening?

CQC doesn't endorse specific tools, but its safer recruitment guidance requires a structured, documented, fair process. An AI screening platform with a full audit trail of questions asked, answers given, and scoring applied is easier to evidence than a pile of CVs marked up in red pen.

Should I ban AI-written CVs?

No — and you couldn't enforce it anyway. Better to reduce how much weight the CV carries in your decision. If the first real assessment is a structured phone call, it matters much less whether the CV was written by the applicant, their cousin, or ChatGPT.

How much does it cost to move away from CV-led hiring?

Lily charges a flat £300 per successful hire on the End to End plan, or £99 for screening only. No subscription, no setup fee — you pay when you hire. See pricing for the full breakdown.

Ready to stop filtering people by CV? Book a demo and see how Lily screens every applicant in under a day.

External reference: Skills for Care — recruitment and retention guidance

The CV is dead. Long live the CV.