AI Wrote the Cover Letter. How Do You Hire Fairly Now?

AI in recruitment: hire for substance, not shine

Picture this. You post a job. Within a day you have sixty applications. Many look brilliant. The cover letters sing. The CVs tick every box. That polish might be the work of AI in recruitment. And that means your old shortcuts for spotting great people stop working.

This is not a panic piece. AI is a tool. Tidying spelling or structure is fine. But AI in recruitment changes the signals you rely on, especially if you read every application yourself. When polish is free, polish stops telling you anything. So let’s focus on what still reveals the truth.

AI in recruitment: what’s actually happening

Free tools create slick, made-to-measure cover letters in seconds. One click. Job done. Candidates can send fifty in the time it used to take to write one. That swells the pile on your desk.

Two things follow.

• Volume goes up. You spend longer sorting, even though each application looks better.
• Old signals go quiet. A neat cover letter once hinted at care and interest. Now it often means the candidate used a chatbot. It tells you nothing about skill or fit.

So here is the real question for recruitment week: in an age of AI in recruitment, what still shows you who can do the job? The answer is the same as it has always been. Structure.

Short answer

Stop grading the shine. Test the work. Use clear criteria, evidence-based questions, practical tasks, and structured interviews.

The answer is structure, not suspicion

Resist the urge to play AI police. Detection tools are unreliable, and they can unfairly flag people who use assistive tech or write in a certain way. Chasing AI in recruitment is a distraction. Build a process that tests substance.

Here is what works in practice.

1) Know exactly what you’re hiring for

You cannot select fairly for a role you have not defined. List the handful of things that truly matter: skills, behaviours, must-haves, nice-to-haves. Start with a proper job description. If yours is fuzzy or copied from 2019, read our guide: How to write a job description. Clarity gives you a fair yardstick to use across AI in recruitment.

2) Score everyone against the same criteria

Decide your criteria up front. Write them down. Score every candidate against the same list. A dazzling letter wins zero points if “dazzling letter” is not on the list. This levels the field in AI in recruitment and keeps bias in check.

3) Ask for evidence, not adjectives

Anyone can say “I am motivated and a team player.” Ask for proof. Use prompts that force detail:

• “Tell us about a time you calmed an unhappy customer.”
• “Walk us through how you would complete this task.” Real examples have texture. Generic AI in recruitment prose does not.

4) Add a short, job-related task

A 20-minute exercise that mirrors the real job beats ten shiny cover letters. For a customer role, draft a reply to a tricky email. For a practical role, a short trial. You learn what matters: what they can do. This is your anchor in AI in recruitment.

5) Structure your interviews

Ask the same core questions in the same way. Take notes and score against your criteria. Structured interviews predict performance far better than a loose chat. In a free-form chat, the confident win. In a structured one, the capable win. This keeps your process solid in AI in recruitment.

Fairness, discrimination risk, and data rules you must respect

Your equality duties hold firm. Do not discriminate, directly or indirectly, on protected characteristics. Be extra careful if you use any AI in recruitment tool to sift or rank candidates. You are responsible for outcomes. Tools can hide bias. For example, silent downgrades for career gaps or non-standard CVs. That is your risk, not the vendor’s.

There is a data protection angle too. Under UK GDPR, people have protections around solely automated decisions that have legal or similarly significant effects. Rejecting someone for a job can be one of those effects. In short:

• Do not let a machine make the final hiring call on its own.
• Keep a human meaningfully in the loop.
• Be open with candidates about how you assess them.
• Only collect and keep the data you need.

Expert quotes you can rely on:

• ACAS: “Employers must avoid discrimination at all stages of recruitment.” Source: ACAS guidance on discrimination.
• ICO: “You must not make solely automated decisions that have legal or similarly significant effects unless certain conditions are met.” Source: ICO guidance on automated decision-making.

If you use AI in recruitment, you can still keep tech in the toolkit. You stay in charge, you can explain your choice, and your notes back you up. Structure helps you do that.

Quick RAG check on common risks

• Red: Solely automated rejections without human review
• Red: Criteria that indirectly screen out protected groups
• Amber: Untested AI screening settings or prompts
• Green: Structured interviews and work-sample tasks with clear scoring

October focus: fairness for Black History Month

A brief truth. Unstructured, gut-feel hiring is where bias lives. “Culture fit” can mask “feels like me.” Polished CVs favour those with time, tools, and confidence, which are not evenly spread. AI in recruitment can widen that gap if you let it.

The fix is the same one we have covered:

• Clear criteria
• Consistent scoring
• Evidence over adjectives
• Structured interviews That is fair hiring. It is also good business.

Your quick hiring checklist

• Define the role with clear must-haves before you advertise.
• Set scoring criteria up front and use one list for everyone.
• Ask for specific, real examples, not claims.
• Add a short, practical task that mirrors the job.
• Structure interviews and keep notes linked to criteria.
• If you use any AI in recruitment tool, keep a human in charge and review outputs for hidden bias.

Useful sources and tools

• ACAS hiring guidance: Avoiding discrimination in recruitment
• ICO guidance: Automated decision-making and profiling
• CIPD resource: Resourcing and talent planning for current UK trends you can cite in your board papers

Need a hand putting this into practice?

If your managers need confident, fair interviewing skills, our training covers bias, structured interviews, and practical scoring. If the admin is tripping you up, our preferred HR software keeps applications and records tidy. If hiring is one of many things to sort before next year’s changes, Ready for 2027 packages contracts, handbook, eight core policies, manager training, and a year of support. You can also see recent media commentary on AI in recruitment on our press page.

Hire for substance, not shine. AI in recruitment can write the letter. It cannot do the job.

Kettle On, Standards Up.

FAQs

• Is it legal to use AI in recruitment to screen candidates in the UK? Yes, but you must avoid discrimination and you must not make solely automated decisions with legal or significant effects without meeting strict UK GDPR conditions. Keep a human in the loop.

• Can I reject a candidate for using AI in recruitment to write their cover letter? Using a tool to tidy writing is not misconduct. Focus on capability. Test skills and evidence. Reject based on fair criteria, not tool use.

• Are AI detection tools reliable? No. They throw false positives and can disadvantage certain groups. Do not rely on them to police AI in recruitment. Use structured selection instead.

• What is the best way to spot real ability early? Ask for short, specific examples linked to your criteria, and set a simple work-sample task. This is the strongest signal in AI in recruitment.

• Do I need to tell candidates if I use AI in recruitment? Be transparent about your process, what tools you use, and how decisions are made. Link to your privacy notice and explain human review.

• What interview format works best now? Structured interviews with standard questions and scoring. Pair with a work sample. Keep notes tied to the criteria.