AI Laws, Regulations & Guidelines Tracker
AI Laws, Regulations &
Guidelines Tracker
The definitive reference tracking every AI governance instrument across every jurisdiction — with enforcement status, penalties, binding nature, and key obligations.
Global AI Governance Map
AI Governance Instruments Tracker
| Country / Issuer | Year | Type | Instrument | Binding? | Status | Approach | Effective Date | Max Penalty | ||
|---|---|---|---|---|---|---|---|---|---|---|
Regulatory Landscape Analysis
EU Model: Risk-Tiered Binding Legislation
The EU AI Act (2024) is the world’s first comprehensive binding AI law. Risk tiers — Unacceptable (banned), High-risk, Limited, Minimal — create a scalable compliance architecture with extraterritorial reach. Penalties reach 7% of global revenue. Full enforcement expected late 2027 under the Digital Omnibus revision.
US Model: Federal Deregulation + State Activism
No federal AI law. Trump EO (Dec 2025) seeks to preempt state AI laws. Colorado (2026), California, and Texas passed comprehensive state laws. All 50 states introduced AI legislation in 2025. The result: a patchwork compliance regime where state law governs unless federally preempted.
Asia-Pacific: Innovation-Led with Binding Instruments Emerging
South Korea’s AI Framework Act (in force Jan 2026) is APAC’s first comprehensive binding law. India’s Guidelines (Nov 2025) are principle-based soft law with sectoral hard law overlays (RBI FREE-AI). China deploys rapid targeted binding regulations for specific harms. Japan uses voluntary frameworks with name-and-shame enforcement.
Global South: Strategy Phase Moving to Binding
African Union Continental Strategy (2024) sets pan-African vision. Brazil’s AI Bill passed Senate but awaits House. Kenya, Nigeria, South Africa in draft strategy phase. Kazakhstan enacted Central Asia’s most comprehensive AI code. Council of Europe Convention (2024) — world’s first binding international AI treaty — anchors human rights across participating states.
Penalty Landscape: From None to 7% Global Revenue
- EU AI Act: Up to €35M or 7% global revenue
- EU GPAI: Up to 3% global revenue
- Canada AIDA (proposed): CAD $25M or 3%
- Italy AI Law: Criminal + up to 5 years imprisonment
- Brazil AI Bill: Up to BRL 50M or 2%
- South Africa POPIA: ZAR 10M or 10 years imprisonment
- China GenAI: Fines + business suspension + criminal liability
GRC Audit Implications
- ISO 42001:2023 AIMS maps to most national frameworks — priority certification for global organisations
- Third-party AI audits: voluntary today; mandatory under EU (2025), India (medium-term), South Korea (ongoing)
- Human-in-the-loop is now universal for high-impact AI across EU, India, South Korea, US states
- Content provenance and watermarking: binding in China (2025), required under EU AI Act, US state laws
- Deepfake controls: binding in China, Italy, California; expected in EU via AI Act enforcement
