Originally published on AI Skills Academy.
An employee pastes a customer email into an AI assistant. A manager uses AI to summarise job applications. A colleague shares a report containing convincing statistics that nobody has checked.
Each task starts with a reasonable intention: save time and get the work done.
But does everyone understand what information they can share, when an answer needs checking, and which decisions require human judgement?
Buying Copilot, ChatGPT or Gemini gives people access to powerful tools. Helping them use those tools responsibly takes something more: practical AI literacy.
And there is a growing compliance reason to take it seriously.
AI literacy already has a place in EU law
Article 4 of the EU AI Act introduced AI literacy obligations for organisations providing or deploying AI systems within its scope. Those obligations began applying in February 2025.
Following amendments in July 2026, organisations must take measures to support staff AI literacy, considering people’s knowledge, experience and the context in which they use AI. The law does not require employers to guarantee a particular level of individual competence.
It also does not prescribe a specific course, certificate or assessment score. The European Commission says organisations can maintain internal records of training and other guidance initiatives. Source: European Commission
For employers, the practical message is clear: consider what your people need to understand, provide appropriate support, and keep track of what you have done.
UK organisations have reasons to act too
The EU AI Act does not automatically apply to every UK business. Its relevance depends on the organisation’s activities and the Act’s territorial scope.
However, UK organisations already have obligations when using personal data in AI systems. The Information Commissioner’s Office publishes guidance on applying UK data protection principles to AI, explaining decisions and assessing risks to individuals. Source: ICO
For employees, those issues become everyday questions:
- Can I put this customer information into an AI tool?
- Is this application approved for work?
- Could this output be inaccurate or unfair?
- Who needs to review it before it influences a decision?
- What should I do if something goes wrong?
An AI policy helps establish the rules. Training helps people apply them when the answer is less obvious.
Confidence and competence can look very similar
Someone may use AI every day and still misunderstand its limitations.
They might write excellent prompts but overlook confidential information. They might produce polished documents but fail to check sources. They might assume that a human approving an AI recommendation is enough, without properly examining it.
A useful learning programme makes these situations visible. It gives people opportunities to practise judgement as well as prompting.
That means helping employees recognise when AI is useful, when to question an answer, and when to involve someone else.
Start with the work your people actually do
A marketing team, an HR department and a finance function will encounter different AI risks. Their training should reflect those differences.
A practical starting point is to:
- Understand current capability. Find out what employees know and where they need support.
- Connect learning to real tasks. Use situations people recognise from their working day.
- Make your policies usable. Explain approved tools, information handling and escalation routes.
- Track learning and follow up. Record participation and revisit gaps as tools and responsibilities change.
This is a sensible approach to workforce development. The exact legal requirements will depend on your organisation and its use of AI.
How AI Skills Academy can help
AI Skills Academy brings assessment, targeted learning and progress tracking together, so you can build a programme around your workforce’s needs.
You can identify skills gaps, help employees develop practical capability, and give managers a clearer view of progress. Business plans add team dashboards, skills heatmaps and manager reporting. Organisation plans allow training to incorporate your own policies and content.
These features can support the learning and reporting elements of a wider AI governance programme. The Academy does not certify legal compliance, and training alone cannot fulfil every obligation relating to AI.
What it can do is help you turn an intention—“we need our people to use AI responsibly”—into a structured programme you can manage.
Explore AI Skills Academy plans.
Find out where your team stands
You do not need to begin with a company-wide rollout.
Start with a small group. Understand their strengths, identify the gaps, and use that evidence to decide what comes next.
AI Skills Academy offers a free AI Capability Assessment for up to 10 employees, with an organisation-level report covering skill scores, heatmaps and an executive summary.
It is a practical first step towards a workforce that can use AI with greater confidence and better judgement.
Start your free team assessment.
Regulatory information checked on 14 October 2026. This article provides general information, not legal advice. Check which obligations apply to your organisation before making compliance claims.