AI skills employers actually want, and how to show them

Learn in your browser · Chapter 01 of 06

Before you write anything, it helps to see this the way the person reading your CV does: the phrase "AI skills" has become so common that on its own, it has stopped meaning anything specific.

Plenty of CVs now carry a line like "strong AI skills" or "proficient with AI tools", usually near the other soft-skills claims. To the person reading it, that line is close to meaningless, because it does not say which tool, for which task, checked in which way, with what result. It reads the same whether the person behind it used an AI writing tool twice for a school project or used one every week to genuinely change how they worked.

This varies a lot by role. A retail or admin role might genuinely value someone who can use a free AI writing tool to draft routine messages and knows to check them before sending. A data-facing role might value someone who can use an AI tool inside a spreadsheet to speed up analysis and knows how to spot when it has made an error. A software role might value someone comfortable using an AI coding assistant as part of writing and reviewing code, not as a replacement for understanding it. None of these are captured by writing "AI skills" as a single line, and a specific example for your own role will always land better than a general claim borrowed from somewhere else.

It is also worth being honest about what this workbook is not. It is not a technical AI-engineering course, and the specific skills a machine learning role needs (built and evaluated properly, not skimmed from a buzzword list) are a different, deeper subject than a general job-seeker's AI fluency. If that is genuinely your target role, treat this workbook as the honesty and communication layer on top of the technical depth you will need elsewhere, not a substitute for it.

Learn in your browser · Chapter 02 of 06

You cannot write a specific example if you have not actually located one. This chapter is a short, honest audit of your own recent work.

Most people who have used AI tools at all have used them more than they would think to mention on a CV, because the use felt too ordinary to count as a "skill". That is exactly the material this chapter is looking for: not a headline achievement, a genuine, specific instance of using an AI tool on real work and getting a real result.

Scroll sideways to see every column.

If your honest audit turns up nothing at all yet, that is useful information too, not a failure. It means the next honest step is to actually try an AI tool on a real, small piece of your current work before your next application, rather than writing a claim you cannot back up. Chapter 5 covers exactly why that shortcut is worth avoiding.

Learn in your browser · Chapter 03 of 06

A genuine example only helps you if it is written down in a way that reads as specific rather than inflated. This chapter gives you the formula and the mistakes to avoid.

A CV bullet about AI use works best when it states four things plainly: the task, the tool, what you personally checked, and the real result. Leave any one of those out and the line drifts back towards a vague claim.

Scroll sideways to see every column.

The same formula works whether the line is a single CV bullet, a line in a cover letter, or a sentence in a LinkedIn summary. Task, tool, check, real result. If you cannot fill in all four honestly, the example is not ready to use yet.

Learn in your browser · Chapter 04 of 06

The question "tell me about a time you used AI" is increasingly common, and most people answer it badly not because they lack an example, but because they answer in generalities instead of walking through one specific moment.

A written CV bullet and a spoken interview answer need the same honest content, but the spoken answer needs a clear shape so it does not wander. A simple structure works well: what the task was, what you actually did with the AI tool, what you personally checked or changed, and what the real result was, including anything you would do differently next time.

Notice what the strong answer does that the weak one does not: it names a specific, ordinary task, admits the tool got something wrong at least once, states exactly what the speaker personally checked, and gives an honest, roughly-estimated result rather than an invented precise one. None of that requires an impressive-sounding project. It requires one real, specific moment, told honestly.

Scroll sideways to see every column.

Learn in your browser · Chapter 05 of 06

This is the mistake this whole workbook exists to prevent, so it gets its own chapter rather than a passing warning.

It is tempting to write "proficient with AI tools" or describe a bigger, more impressive-sounding AI project than you actually delivered, especially when a job advert lists AI skills as desirable and you are worried about being screened out. The problem is not that this is unethical in some abstract sense, though it is; the practical problem is that it almost always gets tested, directly, in the interview.

The damage from being caught out is rarely limited to that one answer. Once an interviewer notices one claim did not hold up under a direct, reasonable question, they tend to listen more sceptically to everything else you say for the rest of the interview, including the parts that were completely true. A single overstated line can cost you credibility on claims you never needed to inflate in the first place.

Scroll sideways to see every column.

None of this means you should undersell genuine skill, and Chapters 2 and 3 exist precisely so you do not have to. The line to hold is simple: describe what you actually did, as specifically and plainly as you can, and never write or say a claim you could not walk through in detail if someone asked you to right now.

Learn in your browser · Chapter 06 of 06

A short recap, and where to go from here.

You can now spot a vague AI-skills buzzword on sight, and you have found two or three genuine examples of your own AI-assisted work worth mentioning. You can write one of those examples as a specific, honest CV bullet using the task, tool, check and result formula, and you can give a spoken answer to "tell me about a time you used AI" that would survive a direct follow-up question, because you built it from something real rather than something borrowed.

This is free educational content published by Mickai LTD, not professional, legal, financial or career-coaching advice, and Mickai accepts no liability for how you use it.

https://trust-agent.ai/exam/ai-skills-for-your-cv-and-interview

Being honest and specific about what an AI tool actually did, and what you personally checked, is the same principle behind Mickai's own approach to trustworthy AI: a human stays in control, and that control is visible, not just claimed.