AI Act: build your compliance roadmap
Understand the European AI regulation, classify your systems by risk level and build a compliance action plan.
- Duration
- 14 h · 2 days
- Level
- Intermediate
- Format
- On site · Remote · Blended
- Max. group size
- 12 participants
Overview
This two-day program gives lawyers, DPOs, compliance officers and project managers the keys to applying the AI Act in their organization. The first day sets the framework — scope, roles, risk levels — and lets participants inventory and classify the AI systems actually in use. The second details the obligations according to the organization's role, the transparency rules and the basics of a documentation audit, then leads to a prioritized roadmap. The AI Act's application timeline is phased and may change (European simplification proposals are under way): the content is updated before each session based on the texts in force.
Who this program is for
- Lawyers, legal advisors and compliance officers
- Data protection officers (DPOs) and CISOs
- Project managers, data scientists and business leads involved in AI deployment
- Executives wishing to frame their compliance approach
Prerequisites
- No mandatory legal or technical prerequisites
- Basic knowledge of GDPR is appreciated
Learning objectives
- Place the AI Act's scope and timeline
- Determine your organization's role: provider, deployer, importer, distributor
- Inventory your AI systems and classify them by risk level
- List the obligations tied to each risk level and each role
- Understand the rules for general-purpose AI models and transparency
- Carry out a first documentation audit of an AI system
- Build a prioritized compliance roadmap
Detailed curriculum
7 modules · 14 h · 2 days
General framework and scope
- History, goals and logic of the regulation: a risk-based approach
- Definition of an AI system and main exclusions
- Territorial scope and application timeline
- Relationship with GDPR, product safety and other legislation
- Fundamental rights, transparency and non-discrimination
Actors and risk levels
- Provider, deployer, importer, distributor: who is who
- Prohibited practices
- High-risk systems: areas concerned and criteria
- Transparency risks and minimal risk
- When a deployer becomes a provider
Determine the organization's role for several use cases, with justification.
Inventory and classify your AI systems
- Method for inventorying uses, including AI built into off-the-shelf software
- Gathering useful information from business teams and suppliers
- Applying the classification grid
- Borderline cases and trade-offs to document
Build a register from the systems actually used by participants and classify them.
Obligations of providers and deployers
- Requirements for high-risk systems: risk management, data quality, documentation, logging
- Human oversight, accuracy, robustness and cybersecurity
- Deployer obligations: compliant use, oversight, informing individuals
- Fundamental rights impact assessment in the cases provided for
- Staff AI literacy and governance
Transparency and general-purpose AI models
- Disclosure obligations: chatbots, generated content, deepfakes
- Rules applicable to general-purpose AI models
- Documentation expected from model providers and transmission obligations
Auditing and documenting an AI system
- Building the evidence file: purpose, data, performance, limits
- Assessing a supplier: questions and contractual commitments
- Traceability, logs and incident management
- Corrective actions
Audit a fictional AI system from its documentation and identify the gaps.
Roadmap and governance
- Prioritizing actions: risk, effort, regulatory deadline
- Roles and responsibilities: sponsor, lead, committee
- Training and raising awareness among teams
- Tracking indicators and periodic review
- Inspections, penalties and handling audits
Each group builds its organization's roadmap and presents it.
Tools and technologies
Teaching methods
- Legal input illustrated with practical cases
- Sector case studies (HR, finance, healthcare, customer relations)
- Classification workshop on the participants' own AI systems
- Guided construction of the roadmap
Assessment
- Prior self-assessment of knowledge
- Group-graded case studies
- Assessment of learning at the end of the course and presentation of the roadmap
Required equipment
- A room with a projector, or a connection to a virtual classroom
- One computer per participant for the workshops
- The list of AI tools used by the organization (optional, very useful)
Resources provided
- Course materials and AI Act glossary
- AI systems register template
- Roadmap and action tracking table template
- Links to the official texts and the European Commission guidelines
Career outcomes
- Initial register of AI systems classified by risk
- Compliance roadmap with priorities, owners and deadlines
- Ability to engage with providers and challenge their commitments