What do employers actually mean when they ask for AI skills?

Job adverts are increasingly asking for "AI skills" without explaining what that means. One posting asks for "experience with AI tools," another wants someone who can "leverage AI to improve efficiency," a third lists "AI literacy" as a desirable attribute, and a fourth simply says "familiarity with emerging technologies including AI." If you are a working professional trying to decide whether you have those skills, or whether a qualification would help you demonstrate them, the vagueness is genuinely unhelpful.

This article takes four of those real advert phrasings, translates each into what the employer is actually asking for, and maps the requirement onto the six named units of the regulated qualifications available at The Data and AI School of London. The honest conclusion is that most of what employers mean by ai skills is judgement, not tool knowledge, and that is exactly why these qualifications are assessed by portfolio rather than exam.

One distinction matters throughout: this article is about using AI, not building it. You do not need to write code, train a model or understand machine learning mathematics to have genuine, demonstrable ai skills for employees. The qualifications described here are for people who work with AI tools in everyday professional contexts, the same way you work with spreadsheets or email, just more deliberately and more critically.

Four real advert phrasings, translated

1. "Experience with AI tools"

At face value this sounds technical. In practice, most employers using this phrase mean something narrower: have you actually opened and used a large language model, an AI writing assistant or a similar tool in a work context, rather than just reading about one? They want to know you are not starting from zero on day one.

The unit this maps to most directly is The use of AI tools and prompts. That unit requires learners to demonstrate that they can select an appropriate AI tool for a given task, construct a prompt with a clear purpose, evaluate the output critically and refine their approach based on the result. If you have done that in a real workplace context and can evidence it, you have what the employer is asking for. If you are not sure how to build that evidence, the blog post How to write a prompt that actually works at work explains the practical skill in detail and is a useful starting point before you begin your portfolio.

2. "Leverage AI to improve efficiency"

This phrase appears in operations, project management, marketing and administrative roles. It is asking whether you can look at a workflow and identify where an AI tool genuinely saves time or reduces error, rather than adding it for its own sake. That is a process analysis skill, not a software skill.

It maps to two units working together: Introduction to AI, which covers what AI can and cannot do, and The use of AI tools and prompts, which requires the learner to apply a tool within a real or realistic task and evaluate whether it added value. The portfolio evidence for these units asks you to demonstrate that you made a reasoned choice, not just that you clicked a button. That is precisely the judgement the employer is describing.

3. "AI literacy"

This is the broadest of the four phrasings and also the most honest one. Literacy implies understanding, not just operation. An employer who asks for AI literacy wants to know that you understand what AI is doing when it produces an output, that you can spot when it is wrong or misleading, and that you are aware of the organisational and ethical context in which you are using it.

This maps across four of the six units: Introduction to AI, Data and AI systems, Ethical and responsible use of AI in the workplace, and Human oversight when working with AI systems. Those four units together form the NCFE Level 2 Award in Artificial Intelligence (AI) for the Workplace, qualification number 610/7671/5. If an employer is asking for ai literacy at work, a regulated qualification covering those four areas is a direct and evidenceable response to that requirement. You might also find it useful to read Can you put an AI qualification on your CV? before you update your profile, because the question of how to present it matters as much as having it.

4. "Familiarity with emerging technologies including AI"

This phrasing often appears in roles that are not primarily technical but where the organisation is changing. It is a signal that the employer wants someone who is not resistant to change and who will continue learning as the tools evolve. It is less about what you know today and more about whether you have a structured habit of keeping up.

The unit that addresses this directly is Continuous learning in AI for the workplace. That unit asks learners to identify how they will stay current with AI developments relevant to their role and to build a personal plan for doing so. It is one of the two units present in the Certificate but not the Award, and it is arguably the most future-facing unit in the qualification. For many employers, this is the differentiator between a candidate who knows a tool and one who will still be useful when the tool changes.

The six units, named and mapped

The NCFE Level 2 Certificate in Artificial Intelligence (AI) for the Workplace, qualification number 610/7672/7, contains all six units listed below. The NCFE Level 2 Award in Artificial Intelligence (AI) for the Workplace, qualification number 610/7671/5, contains the first four. The Certificate contains the Award: if you complete the Certificate you have completed everything in the Award.

Why judgement, not tool knowledge, is the real skill

Every unit in both qualifications is assessed by portfolio. There is no examination. Learners produce evidence from real or realistic workplace activities, which is then marked internally and externally quality assured by NCFE. The grade is Achieved or Not yet achieved.

This approach is not a shortcut. It reflects something important about what ai skills for employees actually are. A multiple choice question can test whether you know what a large language model is. It cannot test whether you noticed that the AI-generated summary you used in a client report contained a factual error, or whether you flagged a policy concern before your organisation used an AI tool on sensitive data, or whether you adjusted your prompting approach when the output was consistently off-target. Those are the things employers are asking about, and they are the things a portfolio can demonstrate.

Most employers are not asking whether you can build an AI system. They are asking whether you can work alongside one responsibly, critically and effectively. That is a different skill, and it is one that can be learned, evidenced and formally recognised.

This is also why the question of what counts as an AI skill is often misunderstood. Knowing the names of AI products is not an AI skill. Being able to use a tool without evaluating its output is not an AI skill. The regulated units described above set out what genuine competence looks like, and that standard is the same for every learner because it is set by NCFE and regulated by Ofqual.

If your employer has already started thinking about how AI tools are governed internally, it is worth checking their position before you begin building portfolio evidence. The article Does your employer have an AI policy? What to check walks through the questions worth asking, and the answers will shape what you can use as evidence in your portfolio.

Frequently asked questions

What counts as an AI skill?

An AI skill is a demonstrable competence in working with AI tools and systems in a professional context. It includes the ability to select an appropriate tool for a task, construct a prompt that produces useful output, evaluate that output critically rather than accepting it at face value, apply the tool within the constraints of your organisation's policy and relevant regulation, and recognise when human judgement needs to override or correct what the tool has produced.

What ai skills on a cv should not mean is a list of product names. Naming a tool tells an employer very little. Demonstrating that you have used it with judgement, and that you have evidence of having done so through a regulated qualification, tells them considerably more. If you are unsure whether your existing experience is enough to present formally, Can you get a qualification in using AI at work? sets out what a regulated qualification involves and who it is designed for.

Do you need to code to have AI skills?

No. The qualifications described in this article, and the skills most employers are asking for in non-technical roles, do not involve coding at any point. The NCFE Level 2 Award in Artificial Intelligence (AI) for the Workplace (610/7671/5) and the NCFE Level 2 Certificate in Artificial Intelligence (AI) for the Workplace (610/7672/7) have no prerequisite qualifications, no technical entry requirements and no minimum prior knowledge of computing. The minimum age is 14. Both qualifications are delivered wholly online and are self-paced.

Building AI systems, training models and writing the underlying code are separate skills that belong to data science and machine learning engineering roles. The distinction matters because conflating the two leads professionals to believe they cannot engage with AI in a meaningful way unless they learn to programme first. That is not true, and it is one of the reasons why what are ai skills is such a common and understandable question. The skills covered by these qualifications are about working with AI, not creating it.

Which qualification is right for you?

The NCFE Level 2 Award in Artificial Intelligence (AI) for the Workplace (610/7671/5) is 60 guided learning hours and costs 249 pounds. It covers four units: Introduction to AI; The use of AI tools and prompts; Data and AI systems; and Ethical and responsible use of AI in the workplace. It is the right choice if you want a focused, regulated qualification that addresses the core ai literacy at work requirement without the extended commitment of the Certificate.

The NCFE Level 2 Certificate in Artificial Intelligence (AI) for the Workplace (610/7672/7) is 100 guided learning hours and costs 395 pounds. It adds Human oversight when working with AI systems and Continuous learning in AI for the workplace to the four units above. It is the right choice if you want the most complete coverage of what employers are asking for, including the units on oversight and ongoing professional development that are increasingly prominent in job specifications.

Neither qualification carries UCAS points and neither is funded: both are sold commercially. Both are Ofqual regulated, which means the standard is publicly verifiable on the Ofqual register by any employer who wants to check it.

Progression from either qualification leads naturally to Level 2 and Level 3 qualifications in digital and data subjects, giving you a pathway if your ambitions extend further once you have completed your first regulated AI qualification.