Data and AI literacy at work: where to start
Data and AI literacy at work: what it covers, why the EU AI Act makes it a priority, and a practical starting point for your teams.
Data and AI literacy is what lets every employee, not just the specialists, work safely and usefully with data and AI tools: knowing when a generative AI assistant is the right tool, what must never be pasted into it, how to check its answer, and why a messy spreadsheet leads to a wrong decision. This section gathers practical guidance for building it across an organisation, with or without Maketools.
Why now?
Three pressures tend to arrive at the same time:
- The tools are already in use. Chat assistants, AI features inside office suites, CRMs and HR software: many employees use AI before anyone has told them how.
- The law expects it. Article 4 of the EU Artificial Intelligence Act asks organisations that provide or use AI systems to take measures to develop the AI literacy of their staff. It has applied since 2 February 2025 and was amended in July 2026. See Article 4 of the AI Act.
- Mistakes are costly and visible. Personal data pasted into a public tool, an invented figure copied into a board report, a budget based on a table full of duplicates: these are failures of habit, not of technical skill.
Data literacy and AI literacy: two sides of one topic
The two ideas overlap, but they answer different questions:
| Concept | Core question | Typical habits |
|---|---|---|
| Data literacy | Can I trust this data and draw a conclusion from it? | spotting duplicates, missing values, inconsistent formats and outliers |
| AI literacy | Should I hand this task to an AI, and how do I check the result? | giving context, setting constraints, spotting hallucinations and bias, keeping a human accountable |
They meet in practice: an AI model is never better than the data it is given, and an analysis produced with AI must be reviewed like any other result.
Where should you start?
A simple approach that fits in a month (detailed in the 30-day AI literacy plan):
- Take stock. Which AI tools are in use, by whom, and for what? Which sensitive data is involved? This map shapes everything else.
- Set a few clear rules. A one-page policy beats a long document: what AI can be used for, what it must never receive (personal data, decisions with consequences), and the rule that nothing is published without review.
- Get people practising. A rule that is read is soon forgotten; a rule applied to realistic cases sticks. That is what serious games are for: short situations, a decision, immediate feedback.
- Measure and keep records. Who took part, which skills were covered, with what results: you need this to steer the programme and to document the measures you have taken.
What each game trains
Maketools games take 4 to 10 minutes, are played solo, and each one measures named skills that stay the same from one game to the next, so reports remain comparable:
| Game | What it trains |
|---|---|
| AI or Not AI? | Deciding whether a request is a job for AI, a human or a script; protecting personal data; keeping a human accountable; reviewing before publishing |
| Prompt Quest | Writing a prompt: context, constraints, examples, output format; avoiding accountability traps |
| Hallucination Hunter | Spotting hallucinations, detecting bias, knowing when to verify |
| Data Cleaning Rush | Spotting duplicates, missing values, inconsistent formats and outliers |
The catalogue keeps growing: open the catalogue to see everything available. Public games can be played without an account.
Who is this for?
Each audience has its own questions, answered on these pages:
- Learning and development teams: rolling out a programme, plugging it into your LMS, evidencing participation.
- Data and AI leaders, DPOs: driving good practice, reducing risk, tracking progress skill by skill.
- Events and workshops: getting a whole room playing without accounts, with a live leaderboard on the big screen.
Unsure about a term? The data and AI glossary defines the concepts used in this section and in the games.
Try it
The quickest way to judge is to play: open the catalogue, pick a game and start it, no account needed. To adapt a game to your organisation, follow Getting started.