What the AI Act requires of financial institutions
- A map of AI systems and their risk level: credit scoring and insurance pricing are among the high-risk cases.
- Requirements on governance, documentation, human oversight and data quality.
- Measures to build AI skills in teams (Article 4, applicable since February 2025 and clarified in 2026: no level to guarantee, no mandatory certificate).
Our view: AI governance belongs to existing functions
High-risk AI systems fall under frameworks that institutions already master: risk mapping, permanent control, model auditability. AI governance is built within the risk and compliance functions, reinforced by hybrid profiles (regulation × data), rather than in a new, isolated function.
How Tyros Group helps
- Advisory: understand, frame and deploy your AI initiatives: governance, compliance, business adoption.
- Training: raise awareness among Executive Committees, regulated functions and business teams (AI Literacy & Responsible Use, Tyros Academy).
- Executive search: recruit the profiles who will carry this governance: CRO, CCO, Chief Data / AI Officers.