In January 2026 Skills England published Skills England AI foundation skills for work — six foundation skills across three domains that describe what people need in order to use everyday AI tools safely and well. Our twelve skills map onto all six, so you can report your workforce's capability against a recognised national reference point.
AI Skills Academy is independent. Skills England does not accredit, approve or endorse this platform; we publish how our skills map to its benchmark so you can judge coverage for yourself.
Nobody is scored on a quiz. Every skill below is measured from real practice: learners write prompts, judge AI answers and work through workplace scenarios, and an AI coach marks what they produce.
Asking for what you actually need: a clear task, the background that matters, and the shape of the answer you want back.
How we measure it
Learners rewrite weak prompts and answer open-text tasks that the AI coach marks on goal, context and output.
Our skills behind it
Knowing what an assistant is good for in everyday work — drafting, summarising, organising — and when to reach for it.
How we measure it
Suitability exercises and workplace scenarios where learners decide whether and how to use AI for a real task.
Our skills behind it
Chaining steps together and handing repeatable work to an assistant or agent, without needing to write code.
How we measure it
Build-a-workflow exercises and agentic modules where learners sequence and supervise multi-step work.
Our skills behind it
Changing tone, format, settings or your follow-up question when the first answer is not good enough.
How we measure it
Iteration exercises that score the learner's follow-up prompts and their control of format, length and tone.
Our skills behind it
Knowing what can go wrong — confidential data, bias, policy breaches — and deciding when to rely on your own judgement.
How we measure it
Security and responsible-use scenarios graded against your organisation's own AI policy and approved tools.
Our skills behind it
Checking outputs for accuracy, spotting errors and invented facts, and recognising patterns you can act on.
How we measure it
Spot-the-mistake exercises, grounding tasks and a full simulation with a deliberately unreliable assistant.
Our skills behind it
| Domain | Skills England foundation skill | AI Skills Academy skills |
|---|---|---|
| Technical | Writing clear instructions for AI tools | Instruction, Context, Output Control |
| Technical | Using AI tools to support routine processes and tasks | AI Fundamentals, Workflow |
| Non-technical | Using simple software that automates work tasks | Workflow, Agentic AI |
| Non-technical | Adjusting how an AI tool works to get better results | Iteration, Output Control |
| Responsible and ethical | Understanding the risks and consequences of AI use | Security, Responsible AI |
| Responsible and ethical | Analysing information using AI tools | Verification, Grounding, Critical Thinking |
Each person's score for a foundation skill is the weighted average of the Academy skills behind it, on the same 0–100 scale we use everywhere else. We treat a foundation skill as met at 60 or above, and someone meets the full benchmark when all six are met. Managers can see this per person and per team, and download it as evidence.
Run the free capability assessment with a sample of your people, and get their results mapped against all six foundation skills.