Choose a region
🇪🇺 EU AI Act
Understand the EU's risk-based AI law, transparency rules, high-risk systems and key deadlines.
Explore EU AI Act →
🇺🇸 US AI Regulation
Understand the U.S. approach to AI regulation, including federal policy and the growing role of state laws.
Explore US AI Regulation →
The three approaches at a glance
Rules differ a lot by region, and dates have moved several times in 2026. This summary reflects what I could confirm in early October 2026. Always check the official text before making a decision.
- European Union: one comprehensive, risk-based law. The main high-risk obligations were postponed in July 2026 to December 2027 for stand-alone systems and August 2028 for AI built into regulated products. Start with the EU AI Act guide.
- United States: no single federal AI law. Rules come from existing regulators and a growing set of state laws, with an unresolved debate over whether federal law should override them. See US AI Regulation.
- India: non-binding AI Governance Guidelines from November 2025, plus binding rules on personal data and on labelling synthetic content. The lesson The Global AI Regulation Landscape covers it.
Who should read what
- A business serving customers abroad: read the EU and US guides, then the governance lesson.
- A creator or publisher using AI images and text: pay attention to labelling and disclosure rules in each region.
- An everyday user: the practical points are in Privacy & Data When Using AI.
Why the names differ: The European Union has a comprehensive law called the EU AI Act. The United States does not have one single equivalent “US AI Act,” so we use “US AI Regulation” for accuracy.
Sources and further reading
Official starting points for each region covered here.