Course catalog
The seven EU AI Act courses in the catalog: who each one is for, how long it takes, its lessons and knowledge check, the pass mark and the version.
The catalog holds seven courses on the EU AI Act. A learner takes the courses for their role, never the whole catalog: most people need one or two.
Each course is one SCORM 1.2 package with a set of lessons and a knowledge check of 22 questions at the end. The app lists the package as the course’s module. The content is written against the consolidated text of Regulation (EU) 2024/1689 as it stood on 27 July 2026.
| Course | Group | Duration | Pass mark | Version |
|---|---|---|---|---|
| AI literacy and responsible use | Standard | 40 min | 70% | 2026.2 |
| AI governance for leaders | Standard | 60 min | 70% | 2026.2 |
| Building, buying and deploying AI | Standard | 90 min | 75% | 2026.2 |
| Secure AI lifecycle and incident response | Standard | 90 min | 75% | 2026.1 |
| High-risk AI and impact assessments | Conditional | 120 min | 75% | 2026.2 |
| AI in employment and HR | Conditional | 60 min | 75% | 2026.1 |
| Annual AI update 2026 | Annual update | 15 min | 70% | 2026.2 |
Groups
Section titled “Groups”- Standard. Assigned by role. AI literacy and responsible use; AI governance for leaders; Building, buying and deploying AI; Secure AI lifecycle and incident response.
- Conditional. Assigned only when a role or use case triggers it. High-risk AI and impact assessments; AI in employment and HR.
- Annual update. Assigned to people who finished the core training. Annual AI update 2026.
A version is frozen once it is assigned. Content changes ship as a new version, so a learner’s record always points at the exact course they took. Course versions and re-training explains what that means for an audit.
AI literacy and responsible use
Section titled “AI literacy and responsible use”
A practical foundation for recognising AI, using approved tools safely, checking outputs, meeting transparency expectations, and escalating concerns before harm occurs.
- Audience: Everyone who uses or is affected by workplace AI
- Group: Standard
- Duration: about 40 minutes
- Knowledge check: 22 questions, 70% to pass
- Version: 2026.2, updated 13 November 2026
Lessons
- Recognise AI in everyday work (6 min)
- Understand how AI fails (7 min)
- Use AI safely at work (7 min)
- Recognise legal red lines (7 min)
- Be transparent about AI (7 min)
- Pause, check, and report (6 min)
After the course a learner can
- Recognise common AI systems, outputs, and limitations in everyday work.
- Apply the organisation’s minimum controls for data, confidentiality, accuracy, and human review.
- Spot prohibited or transparency-sensitive uses and pause them for specialist review.
- Report errors, harmful outcomes, security concerns, and unexpected behaviour through the right channel.
AI governance for leaders
Section titled “AI governance for leaders”
A leadership operating model for accountable AI: assign roles, maintain an inventory, triage use cases, approve proportionate controls, retain evidence, and respond when the risk changes.
- Audience: Executives, directors, business owners, compliance leads, and AI governance members
- Group: Standard
- Duration: about 60 minutes
- Knowledge check: 22 questions, 70% to pass
- Version: 2026.2, updated 13 November 2026
Lessons
- Turn leadership intent into accountable practice (8 min)
- Identify your role in the AI value chain (9 min)
- Build an inventory leaders can rely on (9 min)
- Design a proportionate AI literacy programme (8 min)
- Route each use case through proportionate risk triage (9 min)
- Retain evidence of controls and decisions (9 min)
- Monitor the system and the governance model (8 min)
After the course a learner can
- Distinguish provider, deployer, importer, distributor, and downstream-provider roles for a real use case.
- Establish ownership, approval gates, and role-based literacy measures across the AI lifecycle.
- Use a proportionate risk triage to direct use cases to the right legal, technical, and business review.
- Maintain evidence that demonstrates decisions, controls, monitoring, training, and corrective action.
Building, buying and deploying AI
Section titled “Building, buying and deploying AI”
A lifecycle playbook for turning an AI idea into a controlled release: define purpose, classify the use, establish value-chain roles, assess suppliers and data, design oversight, test, approve, monitor, and manage change.
- Audience: Product, engineering, data, procurement, legal, security, delivery, and client teams
- Group: Standard
- Duration: about 90 minutes
- Knowledge check: 22 questions, 75% to pass
- Version: 2026.2, updated 13 November 2026
Lessons
- Start with intended purpose and boundaries (10 min)
- Classify the use before committing to it (12 min)
- Map roles and responsibilities across the value chain (12 min)
- Perform evidence-led supplier due diligence (11 min)
- Govern data and knowledge sources (11 min)
- Design human oversight and transparency into the workflow (11 min)
- Evaluate the real system and run a release gate (12 min)
- Monitor, control change, and retire safely (11 min)
After the course a learner can
- Write an intended-purpose statement that supports classification, testing, contracting, and monitoring.
- Apply the AI Act triage sequence and identify when the organisation may become a provider or downstream provider.
- Collect and evaluate supplier, model, data, security, transparency, and performance evidence.
- Design a release gate with measurable acceptance criteria, meaningful human oversight, monitoring, and change control.
Secure AI lifecycle and incident response
Section titled “Secure AI lifecycle and incident response”
Security controls for AI-assisted development and AI-enabled products, covering threat modelling, data and secret protection, secure integration, evaluation, monitoring, containment, evidence, and regulatory escalation.
- Audience: Engineering, security, platform, data, product, operations, incident response, and technical leadership
- Group: Standard
- Duration: about 90 minutes
- Knowledge check: 22 questions, 75% to pass
- Version: 2026.1, updated 30 August 2026
Lessons
- Treat AI security as a system property (10 min)
- Use AI coding tools without leaking or inheriting risk (11 min)
- Threat-model AI-specific attack paths (12 min)
- Build least-privilege AI integrations (11 min)
- Evaluate security in delivery pipelines (11 min)
- Monitor without creating a new data leak (11 min)
- Respond to AI incidents (12 min)
- Exercise, measure, and improve resilience (12 min)
After the course a learner can
- Threat-model AI systems across models, data, retrieval, tools, users, infrastructure, suppliers, and outputs.
- Use AI coding and development assistants without exposing secrets or accepting insecure and unlicensed output.
- Design layered controls for prompt injection, data exfiltration, unsafe tool actions, poisoning, and supply-chain change.
- Respond to AI incidents with containment, evidence preservation, impact assessment, notification routing, and corrective action.
High-risk AI and impact assessments
Section titled “High-risk AI and impact assessments”
An advanced, conditional course on high-risk classification, provider and deployer duties, lifecycle controls, conformity evidence, fundamental-rights assessment, human oversight, monitoring, and serious-incident response.
- Audience: Teams providing, deploying, procuring, assuring, or overseeing potentially high-risk AI systems
- Group: Conditional
- Duration: about 120 minutes
- Knowledge check: 22 questions, 75% to pass
- Version: 2026.2, updated 13 November 2026
Lessons
- Classify high-risk AI accurately (12 min)
- Assign operator roles and detect transitions (14 min)
- Translate Articles 8–15 into system requirements (14 min)
- Operate risk management and data governance (14 min)
- Build quality management and conformity evidence (13 min)
- Operate high-risk systems as a deployer (13 min)
- Assess fundamental-rights impact (14 min)
- Make oversight, monitoring, and incident response work (13 min)
- Run the high-risk readiness gate (13 min)
After the course a learner can
- Classify Annex I and Annex III use cases and document any Article 6(3) exclusion conclusion.
- Distinguish provider, deployer, importer, distributor, and Article 25 role transitions for high-risk systems.
- Map Articles 8–15 requirements into risk, data, documentation, logging, transparency, oversight, performance, and security controls.
- Determine when a fundamental-rights impact assessment is required and integrate it with privacy and broader impact work.
- Operate post-market monitoring, corrective action, and serious-incident escalation with defensible evidence.
AI in employment and HR
Section titled “AI in employment and HR”
A conditional course for AI used around workers and candidates: prohibited practices, high-risk classification, equality and privacy, vendor evidence, meaningful human review, worker information, monitoring, challenge, and remedy.
- Audience: HR, recruitment, people operations, managers, procurement, legal, data protection, equality, and worker-representation teams
- Group: Conditional
- Duration: about 60 minutes
- Knowledge check: 22 questions, 75% to pass
- Version: 2026.1, updated 30 August 2026
Lessons
- Map AI across the employment lifecycle (8 min)
- Stop prohibited workplace practices (9 min)
- Classify employment systems (9 min)
- Assess equality, privacy, and worker impact (8 min)
- Procure HR AI with evidence and contractual control (9 min)
- Keep the human decision genuinely human (9 min)
- Inform, monitor, challenge, and remedy (8 min)
After the course a learner can
- Recognise prohibited workplace emotion inference and Annex III employment high-risk use cases.
- Assess intended purpose, decision influence, profiling, data, equality, privacy, and worker impacts before use.
- Evaluate HR AI suppliers and design human oversight that can genuinely change outcomes.
- Inform workers and candidates appropriately and maintain monitoring, challenge, correction, and incident routes.
Annual AI update 2026
Section titled “Annual AI update 2026”
A concise, date-stamped refresher on the July 2026 AI Act amendment, the start of Article 50 transparency enforcement, revised high-risk dates, new prohibited-practice provisions, and the actions every team should take now.
- Audience: Everyone who completed the core AI training
- Group: Annual update
- Duration: about 15 minutes
- Knowledge check: 22 questions, 70% to pass
- Version: 2026.2, updated 13 November 2026
Lessons
- What changed in July 2026 (4 min)
- Transparency rules are live (4 min)
- Refresh the AI portfolio (4 min)
- Reset your responsible-use habits (3 min)
After the course a learner can
- Explain the 2026 change to Article 4 AI literacy and what remains expected of organisations.
- Apply current Article 50 transparency expectations and recognise the synthetic-content transition rule.
- Recall the revised Annex III and Annex I high-risk application dates and new Article 5 additions.
- Update an AI use, disclosure, inventory, or escalation route when the legal or system facts change.