Article 4 of the EU AI Act has been in force since February 2025.
It requires every organization that deploys or uses AI to ensure that its employees possess sufficient AI knowledge, proportionate to their role and the associated risks.
Not "have access to a training module." Demonstrably. Documentably. Proportionate to the specific AI systems in use and the specific risks those people face in their work.
A generic AI awareness course doesn't satisfy this. Checking a box doesn't satisfy this. What satisfies this is a structured programme, mapped to applicable law, with evidence that the right people learned the right things.
That evidence is also what protects individual professionals when something goes wrong. The question in a disciplinary or liability context is not whether the organisation had a policy — it's whether this person understood what they were working with.
The risk is personal as well as organisational.
A lawyer who submits AI-generated content without verification risks professional misconduct. An HR manager who approves AI-driven shortlists without understanding bias exposure risks discrimination liability. A compliance officer who signs off on an AI system without documented literacy training risks personal accountability when that system fails.
Who this is for
Any organisation deploying AI that needs to demonstrate Art. 4 compliance. Legal, compliance, and HR professionals with direct AI risk exposure. Boards and leadership teams that need to understand AI risk at governance level. Organisations preparing for their first AI audit.
Programme structure
Five modules, adapted to your sector and role profile. From a half-day awareness session for whole teams to a multi-day certification track for AI leads, DPOs, and compliance officers. Blended formats available. Completion certificates serve as documented Art. 4 compliance evidence.
Sector profiles available for
Legal & Compliance, HR & Recruitment, Finance & Insurance, Healthcare, Government, Energy & Critical Infrastructure.
Custom sector profiles developed on request.
Includes PE points for lawyers where applicable.
Investment
Intake: 30 minutes, no obligation.
Programme architecture
The curriculum is built on five fixed modules with sector-specific adaptation layers. The adaptation layer modifies case studies, applicable regulatory references, and learning objectives. The module structure and assessment criteria remain consistent across all sector profiles.
What AI actually is How large language models work, what hallucinations are structurally, how the EU AI Act risk classification system works, and which AI system types trigger which obligations. Calibrated to the specific AI systems used by the participant's organisation.
Liability, professional risk, and legal obligations What happens when AI output isn't verified. Sector-specific liability mapping: professional misconduct (legal), discrimination risk (HR), patient safety (healthcare), administrative law obligations (government). Includes case analysis from actual incidents.
Privacy and data protection in AI context GDPR Art. 22 (automated decision-making), Art. 35 (DPIA requirements), processor agreements, and what data protection obligations apply specifically when using third-party AI tools. Scope includes CLOUD Act implications for US-hosted AI services.
AI policy and internal governance What a functional AI policy contains. Approved tool lists, verification procedures, transparency obligations toward clients and data subjects, incident response. How to implement and make this demonstrable to regulators.
Audit readiness and evidence documentation Decision Provenance Bundle construction, training register maintenance as Art. 4 evidence, what regulators and auditors actually request, and how to log AI use in a way that survives scrutiny.
Each level (N1 awareness / N2 operational / N3 strategic) has defined learning outcomes mapped to Art. 4 AI Act, GDPR Arts. 5, 28, 35, NIS2 Art. 20, and DORA Art. 33. Completion certificates include the applicable regulatory mapping as documentation for internal audit use.
Module 5 includes a live Sentinel Compliance Scan of the participant's own organisation as a practical exercise. Findings feed directly into the Gap Analysis component of the module.
Start with a 30-minute intake.
We map which AI systems your organisation uses, which obligations apply, and which programme format fits your team. No commitment required.