AI Governance: The Global Landscape
How different countries and blocs are approaching the regulation of artificial intelligence.
Three Models of AI Governance
Global AI governance is fragmenting into distinct approaches. The EU has taken a risk-based regulatory approach with the AI Act (2024), classifying AI systems by risk level and imposing strict requirements on 'high-risk' applications (hiring, law enforcement, critical infrastructure). It bans certain practices entirely (social scoring, most real-time biometric surveillance).
The US has favored a lighter touch — executive orders, voluntary industry commitments, and sector-specific regulation rather than comprehensive legislation. China has moved quickly on targeted rules — regulating deepfakes, recommendation algorithms, and generative AI — while maintaining state access to AI capabilities for surveillance and social control. The UK has positioned itself as a 'pro-innovation' alternative to EU regulation.