Article 4 AI literacy
Show the measures your organisation takes to support AI literacy.
Identify what each person selects, builds, operates, reviews or governs; close the resulting knowledge gaps through structured learning; then connect approved purpose, limitations, checks and escalation controls to the AI systems they actually touch. Retain the evidence needed for management, customers and reviewers.
Product demonstration · 3 minutes 3 seconds
AI literacy and training assurance
Explore role-based pathways, assessed courses, system briefings, learner evidence, renewals and management reporting using hypothetical data. All organisations, people and systems shown were created for demonstration, and no client data appears.
Watch on its own page with the full transcriptRead the video transcript
AI literacy is not a single course that everyone completes once. People need knowledge and practical capability that reflect their role, experience, technical understanding and the AI systems they use. EU AI Fit turns that principle into a managed programme. The training dashboard shows learner coverage, current evidence, outstanding work, overdue records and the latest assessment results. Managers can see where attention is needed without searching through emails or spreadsheets.
The starting point is role context. In this fictional example, Cedar Vale records that Maya Thompson governs and approves AI, Liam Chen operates a decision-support system and provides human oversight, Beatrice Okafor makes procurement decisions and Noah Williams maintains technical controls. Their pathways are approved against those responsibilities and scheduled for review. A campaign then assigns the appropriate controlled courses.
The same foundation may apply to everyone while governance, effective oversight and technical lifecycle learning are targeted to the people who perform those functions. The campaign shows what is current, in progress, required or overdue, with reminders and evidence export available to the programme manager. Completion is not treated as attendance alone. A learner record can retain the curriculum version, baseline result, final assessment, learning gain, acknowledgement and renewal date.
It also shows outstanding requirements so an 88% pass on the foundation course does not hide a separate governance course that still needs completion. General literacy is only part of competence. People using a named AI system also need operating direction for that system. Here an approved briefing explains the permitted purpose, prohibited uses, limitations, checks, override authority and escalation route for the fictional delivery exception decision support tool.
Each assignment records who acknowledged which version. If the controlled system facts change, the briefing is marked for review before further reliance. For the learner, My Learning brings role-based courses, system briefings, due dates and current evidence into one clear queue. For management, a frozen training assurance report records profile coverage, current and outstanding learning, overdue items, assessment outcomes, measured learning gain and system briefing coverage.
Gaps remain visible and can be assigned rather than concealed by an overall completion percentage. The result is a continuing evidence trail from role analysis to learning, assessment, system competence and renewal. Every organisation, person and scenario shown here is fictional. EU AI Fit supports AI literacy and training assurance. It does not certify competence or replace an organisation's responsibility to judge what its people need.
What people learn
The current course syllabus
Start with shared foundations, then choose depth for the responsibilities each person holds. The course catalogue below comes from the same controlled curriculum used in the workspace. Time estimates are planning guides, not minimum attendance rules or promises of competence.
EU AI Act foundationPeople who select, approve, govern, operate or support AI systems in an organisation · Approximately 120 minutes
Understand how the EU AI Act is structured, identify regulatory roles and risk categories, and connect the applicable duties to real organisational decisions.
Purpose, scope and application
Understand why the Act exists, when it reaches an organisation and why scope must be analysed before risk classification.
AI systems, models and intended purpose
Distinguish the regulated object from its components and use the intended-purpose record rather than marketing terminology.
Regulatory actors and the value chain
Identify who performs each regulated role and recognise when modification or rebranding can transfer provider obligations.
Risk categories, prohibitions and exceptions
Navigate the Act's different regulatory routes without reducing them to a misleading four-level pyramid.
High-risk lifecycle duties
Connect the principal high-risk requirements to the actors who must design, supply, operate and monitor the controls.
Transparency and GPAI
Separate system-level transparency duties from model-level GPAI duties and understand the downstream information chain.
Fundamental rights and connected law
Recognise that AI Act analysis sits alongside data protection, equality, employment, consumer, safety and sector-specific duties.
Implementation, evidence and change
Translate the Act into a controlled governance system that remains accurate as uses, suppliers, law and evidence change.
Assessment threshold: 80%. Completion follows the course’s module assessment and acknowledgement requirements, then its final assessment. A course result does not approve an AI system or establish legal compliance.
Safe and responsible AI useEveryone using AI in their work · Approximately 60 minutes
Apply approved-use, information, verification, fairness, human-review and escalation controls in everyday work.
Work within the approved boundary
Match every AI use to an approved system, purpose and accountable owner.
Protect information and supplier boundaries
Use only information and configurations approved for the AI service.
Verify outputs and communicate honestly
Check material output and apply required disclosure before it affects people.
Consider people, fairness and accessibility
Recognise when an apparently useful AI output could disadvantage or exclude people.
Make human review meaningful
Challenge AI-supported conclusions and retain responsibility for the decision.
Recognise and escalate unsafe use
Contain errors, data exposure, bias and unexpected behaviour through the controlled route.
Assessment threshold: 80%. Completion follows the course’s module assessment and acknowledgement requirements, then its final assessment. A course result does not approve an AI system or establish legal compliance.
AI governance and accountabilityManagers, decision-makers, compliance and governance teams · Approximately 60 minutes
Make proportionate, evidence-based governance decisions and keep accountability effective.
Set accountable roles and decisions
Keep business accountability with named people throughout the AI lifecycle.
Classify and document the use
Use purpose, role, affected people and context to drive proportionate controls.
Plan oversight and affected-person safeguards
Make review, transparency and challenge routes usable in practice.
Monitor, respond and evidence
Use performance, incidents and change signals to keep the decision current.
Assessment threshold: 80%. Completion follows the course’s module assessment and acknowledgement requirements, then its final assessment. A course result does not approve an AI system or establish legal compliance.
System operation and human oversightAI system operators and people responsible for human oversight · Approximately 60 minutes
Interpret, challenge, intervene and escalate AI-supported decisions effectively.
Understand the system and decision boundary
Know the intended purpose, inputs, outputs, limitations and consequences of intervention.
Resist automation bias
Actively test AI-supported conclusions instead of rubber-stamping them.
Intervene, override and use fallback routes
Make stop, override and alternative-review controls practical.
Record and escalate oversight signals
Turn interventions, errors and near misses into governance evidence.
Assessment threshold: 80%. Completion follows the course’s module assessment and acknowledgement requirements, then its final assessment. A course result does not approve an AI system or establish legal compliance.
AI lifecycle and technical controlsDevelopers, integrators, technical owners and assurance specialists · Approximately 60 minutes
Control data, evaluation, security, change, monitoring and traceability across the AI lifecycle.
Control data and model limitations
Document provenance, representativeness, quality and the limits of the selected model.
Evaluate performance, robustness and fairness
Test against the intended context, foreseeable misuse and consequence of failure.
Secure the system and control change
Protect the AI supply chain, interfaces, logs and releases.
Monitor and support incident response
Connect technical signals to operational thresholds, investigation and governance decisions.
Assessment threshold: 80%. Completion follows the course’s module assessment and acknowledgement requirements, then its final assessment. A course result does not approve an AI system or establish legal compliance.
From a learning result to operating practice
In this scenario, a procurement lead can explain why an unsupported supplier claim needs evidence, but still needs the organisation’s own approval and escalation rules. The manager combines the relevant course with an approved system briefing, retains the assessment and acknowledgement record, and sets a renewal or change-triggered review. A completion percentage should never hide an outstanding briefing.
Business outcomes
Turn readiness work into reviewable evidence
Find the knowledge gap
Map every responsibility a person has for selecting, operating, overseeing, building or governing AI, then approve an explainable learning pathway.
Learn and demonstrate
Use source-linked, versioned courses with unresolved cases, module assessments, key-learning summaries and acknowledged sign-off before an unseen final assessment.
Brief every system team
Generate a system operating briefing from controlled inventory and risk facts, require human approval, assign it to named people and retain version-specific acknowledgements.
How it works
A controlled path to a usable record
Profile
Identify everyone who selects, implements, uses, reviews, manages or governs AI and capture the responsibilities that shape their learning.
Learn
Approve the recommended pathway, complete practical modules and use assessment feedback to close remaining knowledge gaps.
Apply
Approve system-specific boundaries and checks, assign them to operators and owners, and trigger review when the underlying system record changes.
Built for scrutiny
A learning record, not a generic certificate
EU AI Fit helps organisations demonstrate a reasoned, context-specific approach. It does not claim that one course or certificate guarantees compliance.
Understand the training assurance recordA three-hour common pathway combining EU AI Act foundations with safe and responsible use
Explainable multi-responsibility pathways for decision-makers, operators, oversight, technical and governance teams
Managed training campaigns with role targeting, bulk assignment, branded invitations, reminders and live completion reporting
Unresolved case files and unseen transfer assessments with full module coverage and targeted remediation
Human-approved system briefings generated from live purpose, risk, oversight and supplier facts
Frozen, fingerprinted management reports with PDF, evidence CSV and individual learner timelines
Continue your readiness journey
Useful next steps
Connect learning to role and risk decisions
See how reviewed operator roles, intended purpose and risk classification provide context for system-specific briefings.
Explore this nextUnderstand Article 4
Read the practical EU AI Act guide to proportionate, role-aware AI literacy measures.
Explore this nextPlan the implementation
Define the people, pathways, system briefings, assessment evidence and renewal model before assigning courses.
Explore this nextSee the complete platform
Explore how training records connect to AI inventory, obligations, evidence and governance actions.
Explore this nextPlan a supported pilot
Compare the available readiness and managed-service options for your organisation.
Explore this nextConnect what people learn to the AI systems they actually operate.
Use the pilot to deliver controlled role-based learning, system-specific operating direction and evidence that remains connected to governance change.
Discuss an AI literacy pilot