The leader track 🤝 · for C-level, boards & AI leads · 6 x 45 min · no jargon
What the law requires, when it reaches you, and how the voluntary frameworks turn it into something you can run - each session ends with questions to ask your AI, legal, and data teams.
Why AI governance now + the Act's reach
The EU AI Act reaches a Singapore company when its AI output is used in the EU. Why governance is a board topic, and the map of what is coming.
🪜Leader 2 · easyThe risk tiers
Prohibited, high-risk, limited, minimal - plus the GPAI layer. The one mental model that sorts any AI system and sets what you owe.
🎭Leader 3 · mediumWho must do what + the timeline
Provider vs deployer vs importer - who carries the load - and the dates the obligations actually bite (with the live amendment).
⚖️Leader 4 · mediumPenalties + enforcement
Up to EUR 35M or 7% of global turnover for banned uses. The three fine tiers, the AI Office, and national enforcement.
🧩Leader 5 · decidingNIST AI RMF · ISO 42001 · SG framework
The Act says what you must do; NIST says how to think about risk; ISO 42001 lets you run and prove it. How they fit together.
🏛️Leader 6 · decidingBuilding an AI governance program
The roadmap, the roles, and how AI governance ties to your data governance and your evals - one operating system, not three.
The practitioner track 🛠️ · for AI/ML, risk & compliance · 8 x 45 min · Vantage case
The operational half: how to actually classify, comply, and prove it. Every session moves Vantage, a Singapore firm shipping an AI hiring tool into the EU, closer to defensible governance.
Classify a system - the checker
A live risk-tier classifier: pick what an AI system does and see its tier, the obligations that attach, and the article behind it.
🛡️Practitioner 2 · mediumHigh-risk obligations
Risk management, data governance and bias, technical docs, logging, human oversight, accuracy - the seven a high-risk provider must meet.
💬Practitioner 3 · mediumTransparency obligations
Tell people it is AI, mark synthetic content machine-readably, disclose deepfakes. The Article 50 duties that catch chatbots and GenAI.
🧠Practitioner 4 · hands-onGPAI / foundation models
Baseline documentation, copyright policy, training-data summary - and the systemic-risk tier above 10^25 FLOP with adversarial testing.
🔁Practitioner 5 · hands-onNIST AI RMF hands-on
Govern, Map, Measure, Manage - the voluntary risk method that gives your team a common language and maps onto both the Act and ISO.
📐Practitioner 6 · hands-onISO/IEC 42001
The first certifiable AI management system - policy, roles, impact assessment, Annex A controls - the way to prove governance to a regulator.
📋Practitioner 7 · hardestRisk assessment + conformity
The fundamental rights impact assessment, and the conformity-to-CE-to-registration path a high-risk system runs before market.
🗂️Practitioner 8 · hardestThe AI governance operating model
An AI registry, clear roles, evals as compliance evidence - and running AI governance in one system with GDPR and the PDPA. Vantage graduates.
The running case + the live classifier 🔎
Vantage, and a risk-tier classifier you can poke
The practitioner track follows Vantage, a fictional Singapore firm shipping an AI hiring tool to EU employers - a textbook high-risk system under Annex III(4). Session b1 gives you a live classifier: pick what an AI system does and it returns the EU AI Act tier (prohibited / high-risk / limited / minimal / GPAI), the obligations that attach, and the article behind it. It encodes the Article 5 / 6 / 50 / Chapter V logic (educational, not legal advice).
Choose your path 🗺️
Two doors, one course. Leaders get the law and the frameworks; practitioners build the program that proves compliance.
The knowledge map 🧠
Both tracks at a glance - hover a session to spotlight its concepts, click any node to jump in.