EU Artificial Intelligence Act
The world's first comprehensive AI law. Risk-tiered approach — prohibited, high-risk, limited-risk, and minimal-risk AI. High-risk systems face conformity assessments, technical documentation, and post-market monitoring.
General-purpose AI models above 10²⁵ FLOPs face additional systemic risk obligations. Fines reach €35M or 7% of global turnover.
Read the EU AI Act →NIST AI Risk Management Framework
Voluntary but widely adopted GRC structure with four core functions: Govern, Map, Measure, and Manage. Structured approach to identifying and mitigating AI risks across the entire lifecycle.
Version 2.0 includes updated governance guidance and expanded implementation resources.
NIST AI RMF →ISO/IEC 42001:2023
International standard for AI Management Systems (AIMS). Certifiable, modelled on ISO 27001. Provides structural backbone for AI GRC programmes.
Demand for certification is rising as the EU AI Act references ISO standards. Particularly adopted in Europe and Asia-Pacific.
ISO/IEC 42001 →UK Pro-Innovation AI Approach
Principles-based, sector-led model. ICO, FCA, CMA, and Ofcom each apply existing powers to AI in their domains. The AI Safety Institute runs frontier model evaluations.
UK AI Policy →Singapore Model AI Governance Framework
MAS and IMDA co-developed one of the world's most detailed voluntary AI governance frameworks for financial services and general use. Widely respected as a balanced, innovation-friendly model.
Singapore AI Framework →Framework Comparison
| Framework | Origin | Type | Scope | Status |
|---|---|---|---|---|
| EU AI Act | EU | Mandatory | Comprehensive risk-tiered AI regulation | Phased enforcement 2024–27 |
| NIST AI RMF | US | Voluntary | Risk management (Govern, Map, Measure, Manage) | v2.0 published |
| ISO/IEC 42001 | Intl | Certifiable | AI Management Systems | Active, demand rising |
| UK AI Approach | UK | Principles | Sector-led, regulator-applied | Evolving |
| Singapore MAIGF | SG | Voluntary | AI governance for financial services | v2 published |
| OECD AI Principles | OECD | Principles | High-level policy alignment | Global reference |
Regulatory Timeline
Prohibited Systems Banned
Social scoring, manipulative AI practices prohibited under EU AI Act.
GPAI Model Obligations
General-purpose AI providers face transparency and documentation duties.
Article 50 — Transparency Live
Chatbot disclosure and AI-generated content labelling now mandatory.
AI-Generated NCII Prohibitions
New prohibitions on AI-generated non-consensual intimate imagery.
High-Risk AI Full Obligations
Full conformity requirements enforced (extended by Digital Omnibus).