AI in Recruitment Was Meant to Help. Candidates Are Increasingly Fed Up

First, the candidate gets a rejection. Not pleasant, but it happens. Then they read further and see something they were definitely not meant to see: instructions for a chatbot. Something like: write a warm, generic rejection email, polite but firm, without giving specific reasons. The recruitment process of the future? More like a digital stumble in smart shoes.

In recent months, examples have gone viral showing that automated recruitment has a second, less impressive side. There was the rejection email in which a company accidentally left in its AI prompt. There was the viral video of Leo Humps, who realised he was talking not to a person but to an AI recruiter, and the system started to glitch in the middle of a question. The video was watched millions of times, and the rejection email that followed reportedly got his name wrong.

Then there is the data. According to Fortune, 38% of candidates dropped out of a recruitment process when it required an interview with AI. Another 12% said they would do the same if such a stage appeared in their process. That is not a minor dislike of new technology. It is a sign that some people are starting to treat AI in recruitment as a red flag.

Is the problem AI, or how it is used?

The technology itself is not the villain of this story. AI can help recruiters organise applications, draft job descriptions, reply to candidates faster or spot the match between a CV and the requirements. In large recruitment drives, where hundreds of applications pour in, automation can be a lifeline against an administrative avalanche.

The problem starts when the tool pretends to be a relationship. A candidate spends time on their CV, a test, an interview and preparation, and gets back a message so generic it could have been sent at the same time to an accountant, a graphic designer and an alpaca trainer. It is even worse when the system runs the conversation without clearly saying so, or behaves as if it does not know whether it is a recruiter, a form or an avatar from the "we no longer have people to talk to people" department.

Candidates often do not know the rules

UK employers, like employers everywhere, are experimenting with AI at different stages of hiring, from screening CVs to scheduling and running first-round interviews. Many, however, do not tell candidates clearly what the tools do.

And that is the heart of the matter. Candidates do not always know whether their CV is assessed by a person, an applicant tracking system, an algorithm or all of them at once. Nor do they know whether they are allowed to use AI to improve their CV, prepare answers or research the company. Without clear rules, recruitment turns into a board game where someone has hidden the instructions under the carpet.

How to tell that AI is part of the process

You cannot always be sure, but there are signs worth noticing. Automated recruitment can show up in different ways:

• a very fast reply after you send your CV, sometimes within minutes;
• a generic email that does not refer to your experience or to a specific conversation;
• a form with behavioural questions and no contact with a recruiter;
• a one-way video interview, where you record answers with nobody on the other side;
• an avatar or synthetic voice that runs the conversation from a rigid script;
• no chance to ask questions about the company, the team, the conditions or the next stages.

Not every one of these means a bad process. Online tests, forms and automatic confirmations can be a normal part of recruitment. But if your entire contact with a company feels like talking to an automated helpline, you have every right to feel you are being judged without a real chance to show who you are.

AI can filter candidates, but it can also scare them off

For employers, the key lesson is this: automation must not destroy the candidate experience. A company may save time at the first stage but lose good candidates if the process is unclear, cold or technically clumsy.

An interview is not just a test of the candidate. It is also the moment when the candidate judges the company. If the first impression comes from a glitching bot, a misspelt name or an email with the prompt still in it, it is hard to argue afterwards that the organisation cares about the quality of its communication.

The point is not to ban AI from HR. It is about sensible limits. AI can help organise the process, but decisions about people should not disappear into a black box. Candidates should know when they are talking to a person, when to a system and what information is being taken into account. UK data protection law already gives you rights here: you can ask an employer what personal data it holds about you, and where a significant decision is made solely by automated means, you can ask for a person to review it.

How to prepare for an AI interview

If the process includes an automated stage, do not panic. Treat it as a particular interview format: less spontaneous, more structured, sometimes clunky, but still something you can get through.

Speak clearly, specifically and in plain language. AI may struggle with irony, tangents or overly casual answers. The STAR method works well: situation, task, action, result. So instead of "I work well under pressure", give a short example: what the problem was, what you did and what the outcome was.

5 practical tips for candidates

• Check who you are talking to. If the invitation mentions an "automated interview", "AI interview" or "video assessment", prepare for an interview without a live recruiter.
• Do not improvise without a plan. Prepare four or five examples from your experience: a difficult situation, teamwork, a problem you solved, a conflict, an achievement.
• Keep it simple and to the point. Automated systems may pick up specific words, a clear structure and a match with the requirements more easily than subtle hints.
• Get the tech right. Camera, microphone, lighting and a stable connection all matter; our guide to video interviews from home covers the basics. If the system freezes, note the time, take a screenshot and tell the company.
• Do not be afraid to ask about the rules. You are entitled to ask whether AI is involved in the assessment, what the next stage looks like and whether a person makes the final decision.

AI in recruitment can help, but only when it does not pretend to be human and does not replace thinking. Candidates do not expect every rejection email to be a literary gem. They expect a minimum of respect, clear rules and a process in which technology does not turn them into rows in a spreadsheet. Because even the best algorithm cannot fix the impression that a company has forgotten there is a person on the other side.

Looking for your next role? Browse current jobs in the UK on Jober.