The questions people (and AI assistants) ask most about AI regulation — with short, sourced answers. Every count is a live figure over our primary-sourced records, and each answer links to the page where you can verify it.
As of Jul 4, 2026, 7 US states have an AI-specific law that is enforceable against businesses today. A further 3 have signed an AI law with a compliance deadline still pending, and 20 are actively moving AI bills or running study committees. The remaining 20 have no comprehensive state AI statute — the federal framework and existing consumer-protection law still apply there. Counts are live over all 51 US jurisdictions (50 states + Washington D.C.).
The EU AI Act entered into force on 1 August 2024 and applies in phases: bans on “unacceptable-risk” AI and AI-literacy duties from 2 February 2025; general-purpose AI (GPAI) model obligations and governance rules from 2 August 2025; general application, including most high-risk system requirements, from 2 August 2026; and rules for high-risk AI embedded in already-regulated products from 2 August 2027.
11 US jurisdictions we track have AI-in-employment provisions in their AI laws or bills — covering things like notice to candidates and bias/impact assessments for automated hiring, promotion, and discipline decisions: Alaska, Delaware, Hawaii, Idaho, Illinois, Massachusetts, Michigan, New Jersey, New Mexico, New York, Washington D.C.. Requirements and enforcement status vary by state, so check each state's page for what actually applies.
Penalties depend on the jurisdiction. Under the EU AI Act, fines reach up to €35M or 7% of global annual turnover for prohibited practices, with lower tiers for other breaches. In the US there is no single federal AI penalty; state AI laws are typically enforced by the state Attorney General — often as an unfair or deceptive trade practice or with a civil penalty per violation — rather than by a dedicated AI regulator. Always confirm the exact figure against the specific statute for your jurisdiction.
There is no single comprehensive federal AI statute in the United States. AI is instead governed by a framework of existing laws (civil-rights, consumer-protection, and sector rules), agency guidance, and executive action — which is why AI regulation is currently being written state by state on top of that federal baseline.
In a growing number of states, yes — where an AI or bot interacts with people, transparency/disclosure duties can require telling the person they are dealing with AI. We track transparency or disclosure provisions across many US jurisdictions, with chatbot- and companion-AI-specific duties emerging in several. Whether a rule binds your specific use depends on the state and context, so check the relevant state page.
Beyond the US, the EU AI Act binds all EU member states, and we track 11 more national regimes: Australia, Brazil, Canada, China, India, Japan, Saudi Arabia, Singapore, South Korea, United Arab Emirates, United Kingdom. Their approaches range from comprehensive AI Acts to sector-specific rules and data-protection-anchored regimes.
You have three options with AI Law Tracker: read the weekly "This Week" digest, poll the developer API’s /v1/changes changelog on a cursor, or subscribe an HMAC-signed webhook to get a push on law.created / law.updated events. A free self-serve API key covers polling; webhooks are on the paid tiers.
Informational only — not legal advice. Verify against each linked primary source before relying on it.