Understand, then question

AI literacy course: understand how AI works well enough to use it and question it

This AI literacy course teaches you how AI tools produce their answers, why they make mistakes, where bias and data risks come from, and how to question an AI system before you rely on it. It suits staff and managers who use AI or are affected by it, and organisations building AI literacy for their people.

For you or your whole team

Start with a useful first step.

7-day free trial. No credit card needed.

You will continue on the Tosheo.ai app to create your account and review the trial terms.

What you will learn

The modules in this course.

How AI produces answers

Explain training data and prediction in plain terms and what that means for reliability.

Recognising failure modes

Spot invented facts, outdated knowledge and overconfident answers.

Understanding bias

Describe how bias enters AI systems and where it causes harm at work.

Protecting data

Know what may and may not go into an AI tool and why settings and contracts matter.

Questioning AI systems

Ask vendors and colleagues the right questions about data, testing, oversight and accountability.

Working within rules

Apply your company's AI policy and understand the EU AI Act's AI literacy duty at a high level.

Example task / fictional source material

Capstone: an AI risk note for one real use case

Choose one AI tool or use case your team uses or is considering. Write a one-page note for colleagues covering how the tool produces its output, two ways it could be wrong, one possible source of bias, what data it should never receive and who checks its output before anyone acts on it. Use only non-confidential information.

Review criterion

Accurate explanation

The note explains how the tool works in plain language without overstating what it can do.

Review criterion

Specific risks

The two failure modes and the bias source are specific to this use case, not generic.

Review criterion

Data boundary

The note names concrete types of data to keep out.

Review criterion

Human oversight

A named role checks output, and the note says what happens when it is wrong.

Copy and adapt

Prompts to try

Replace the parts in square brackets with your own details, and check every output before you use it.

Find the weak points in an answer

Here is an answer an AI tool gave me about [topic]. List each factual claim, say which ones I should verify, and suggest where I could check each one. Do not add new claims. Answer: [paste the AI answer]

Prepare questions for an AI supplier

I am evaluating an AI tool that [what it does] for a [size and type] organisation. Draft eight questions to ask the supplier about training data, bias testing, accuracy, human review and how errors are handled. Explain in one line why each question matters.

Explain bias in a use case

Explain in under 200 words how bias could enter an AI system used for [use case], giving two concrete examples and one practical check a non-technical team could run.

Every lesson

10 lessons you can start today.

Your diagnostic picks the order and skips what you already know. Start any lesson with a 7-day free trial, no credit card needed.

Intro to responsible AIExplain fairness, transparency, privacy and accountability
FoundationRead the free lesson Start this lesson
Ethical AI at workDecide when an AI use crosses a line
FoundationRead the free lesson Start this lesson
Your company's AI policyFollow an acceptable use policy in daily work
FoundationStart this lesson
Writing an AI policy for a small businessDraft a one-page policy and get it reviewed
AdvancedStart this lesson
Data protection and AIApply privacy rules to AI inputs and outputs
PractitionerStart this lesson
Human in the loopDecide where a person must review AI work
PractitionerRead the free lesson Start this lesson
Disclosure and transparencyDecide when to tell people AI was used
PractitionerStart this lesson
AI literacy and the EU AI ActExplain the Article 4 AI-literacy duty for organisations
PractitionerStart this lesson
Evaluating AI vendorsAsk about data use, security and model training
AdvancedStart this lesson
Leading responsible adoptionPlan a rollout with training, policy and review
AdvancedStart this lesson

Who is this AI literacy course for, and who is it not for?

It is for anyone whose work now touches AI: people using chat assistants, managers approving AI tools, and staff whose decisions are informed by AI scores or suggestions. It is also useful for the person asked to organise AI literacy training across a small or mid-size company.

It is not a technical course on building or training models, and it is not legal advice. If you want to code, the optional Code track teaches Python and machine learning in code from zero; if you already work as an engineer and need advanced specialist training, look for a machine learning engineering course. If you need a legal view of your obligations, speak to a qualified adviser.

What will I learn in the AI literacy course?

The course moves from mechanism to judgement. You first learn enough about how AI works to predict its failures, then practise questioning real tools and use cases.

1. How generative AI produces an answer

Training data, patterns and prediction, explained without maths, and why the same question can get different answers.

2. Why AI gets things wrong

Invented facts, out-of-date knowledge, overconfidence and the limits of what a tool can know about your situation.

3. Bias and fairness

How bias enters through data and design, and what it looks like in hiring, lending, customer service and content.

4. Data, privacy and confidentiality

Where your inputs go, what settings and contracts change, and what never belongs in a prompt.

5. Questioning an AI system

The questions to ask a vendor or colleague: what data, what checks, who is accountable and what happens when it is wrong.

6. AI rules and your organisation

Company AI policy, human oversight and a factual overview of the EU AI Act's AI literacy duty.

What does the EU AI Act say about AI literacy?

Article 4 of the EU AI Act (Regulation (EU) 2024/1689) requires providers and deployers of AI systems to take measures to support AI literacy among their staff and others operating or using AI systems on their behalf, taking account of their knowledge, experience, training and the context of use. According to the European Commission, Article 4 has applied since 2 February 2025, and national market surveillance authorities supervise and enforce the rules from August 2026.

The Commission's AI literacy questions and answers say there is no need for a certificate, and that organisations can keep an internal record of training and other initiatives. It suggests building a general understanding of AI, working out whether the organisation is a provider or a deployer, assessing the risks of the systems used and tailoring actions to staff and context. The Commission also reports that the Digital Omnibus on AI amended Article 4 in 2026, so check the current text before relying on any wording.

Tosheo.ai can be one part of an organisation's AI literacy measures, but completing this course does not make anyone or any organisation compliant. Compliance depends on your systems, roles and risks, and is a judgement for your organisation and its advisers.

How does the personalised path adapt to my role?

The diagnostic asks how AI touches your work: do you use it, approve it, or receive its outputs? A recruiter's path spends longer on bias in screening tools, a marketer's on invented claims and copyright questions, and a team lead's on approving tools and setting oversight. Lessons you already know can be skipped after a quick check.

What do I practise in an AI literacy course?

You practise judging AI, not just using it. Tasks ask you to find the error in an answer, predict how a tool might be biased, or write the questions you would put to a supplier. Each is marked against a visible rubric with specific feedback, and you can flag a score for review.

Example task: a fictional 40-person freight company, Northgate Freight, is offered an AI tool that ranks job applicants. You write five questions for the supplier about training data, testing for bias, human review and appeals. The rubric checks that each question is specific, that at least one covers how the tool was tested on groups like Northgate's applicants, and that one asks who can override a ranking.

How long does the AI literacy course take?

Lessons last 10 to 15 minutes. How many you take depends on your diagnostic and your pace, so we do not set a fixed number of hours.

Is there a certificate for AI literacy?

Yes. After a final practical assessment, where you analyse an AI use case and its risks, you receive a recognised certificate. It is not an accredited qualification, and as the Commission notes, the EU AI Act does not require a certificate. Team plan managers can see skill progress and a team report, which some organisations use as part of their internal record.

What does the AI literacy course cost?

It is included in both plans. Individual costs USD 79 a month or USD 790 a year. Team costs USD 189 a month or USD 1,890 a year for up to 3 members; each extra member is USD 49 a month or USD 490 a year, and organisations with 25 or more members get volume pricing. Try it with a 7-day free trial, and no credit card is needed.

Sources and scope

European Commission, AI literacy questions and answers, checked 19 October 2026 ↗

EUR-Lex, Regulation (EU) 2024/1689 (Artificial Intelligence Act), checked 19 October 2026 ↗

Before you decide

A few good questions.

What is an AI literacy course?

An AI literacy course teaches people how AI systems work, where they fail, the risks around bias and data, and how to use and question AI responsibly in their own work. It focuses on understanding and judgement rather than coding.

Does this AI literacy course make my organisation compliant with the EU AI Act?

No. Training can be one of the measures an organisation takes under Article 4, but compliance depends on your systems, roles and risks. Tosheo.ai does not certify compliance.

Does the EU AI Act require an AI literacy certificate?

The European Commission's AI literacy questions and answers say there is no need for a certificate. Organisations can keep an internal record of training and other initiatives.

Is AI literacy the same as knowing how to use ChatGPT?

No. Using a tool is part of it, but AI literacy also covers understanding why tools make mistakes, recognising bias, protecting data and questioning AI systems before relying on them.

Can I try the course for free?

Yes. Every plan starts with a 7-day free trial, and no credit card is needed.

Give your team a path, not another playlist.

Explore role-specific lessons, practical tasks and the manager view.

Start free trial7-day free trial · No credit card needed

AI literacy course: understand how AI works well enough to use it and question it

Start with a useful first step.

The modules in this course.

How AI produces answers

Recognising failure modes

Understanding bias

Protecting data

Questioning AI systems

Working within rules

Capstone: an AI risk note for one real use case

Accurate explanation

Specific risks

Data boundary

Human oversight

Prompts to try

Find the weak points in an answer

Prepare questions for an AI supplier

Explain bias in a use case

10 lessons you can start today.

Who is this AI literacy course for, and who is it not for?

What will I learn in the AI literacy course?

1. How generative AI produces an answer

2. Why AI gets things wrong

3. Bias and fairness

4. Data, privacy and confidentiality

5. Questioning an AI system

6. AI rules and your organisation

What does the EU AI Act say about AI literacy?

How does the personalised path adapt to my role?

What do I practise in an AI literacy course?

How long does the AI literacy course take?

Is there a certificate for AI literacy?

What does the AI literacy course cost?

Sources and scope

A few good questions.

Keep exploring.

Guide: what AI literacy means at work

Plans and pricing

Browse every course

Start with the AI course for beginners

Writing an AI governance framework

Give your team a path, not another playlist.