Global Policy Monitor
Sovereignty // Regulation // Alignment
THE RULES OF
THE GAME
Tracking the global race to regulate Artificial General Intelligence. From the EU's strict safety protocols to the UAE's open-source acceleration.
Sovereignty Index
LEGISLATIVE DOCKET
EU AI Act
Forces immediate audit of all foundation models. Non-compliant US labs risk losing access to the entire EU single market. Penalties up to 7% of global turnover.
Executive Order 14110
Mandates "Red Teaming" reports for models >10^26 FLOPs. Effectively creates a "license to train" for frontier labs (OpenAI, Anthropic, Google).
Interim Measures for GenAI
Strict content control requires alignment with "Socialist Core Values". Creates a bifurcated AI ecosystem: one for China, one for the world.
AI Safety Institute Guidelines
Establishes the UK as the global "neutral arbiter" for safety evaluations. Voluntary compliance now, but likely precursor to binding standards.
UN Global AI Governance
First attempt at a "Global AI Treaty". Likely to fail on enforcement but will set the normative standards for "Equitable Access" to compute.
Live Policy Signals
Global Alignment Index
Global safety protocols are currently lagging behind model capabilities.
Official Sources
- EU Commission AI Office
- NIST AI Safety Institute
- Cyberspace Admin of China