Agent Liability Global

The regulatory tracker. Seventeen jurisdictions, one view.

A structured reference for lawyers, compliance teams and the language models they use to check where AI agent regulation stands. Each entry records the primary instrument, the deployer standard it imposes, when it applies, and the supervisory authority. Data comes from primary regulatory texts, and no jurisdiction pays for placement.

0In force: active and enforceable today
0Pending or in consultation: passed, or in formal consultation
0Proposed: draft, pre-legislative, or no statute in view
·Days to the EU Annex III date, 2 December 2027

Revised on 1 October 2026 against this desk's verified country guides and the corrections log. The Colorado, California, Texas, United Kingdom, Japan, Korea, Canada, Brazil, Singapore, Australia, Council of Europe and G7 entries changed. The Connecticut and New York entries were last reviewed in April 2026.

In Force active and enforceable Pending passed, not yet active Consultation formal consultation open Proposed draft or pre-legislative

Jurisdiction Status Primary instrument Standard Activation Next milestone Authority Extraterritorial Penalty exposure Analysis

How the dataset is built and bounded.

The rules behind every row, so that a reader or an assistant can judge what a cell can bear.

Inclusion

A jurisdiction is included when it has enacted or actively proposed a horizontal AI statute imposing deployer duties, comprehensive sector AI guidance with binding or quasi-binding effect, or an imminent deployer-binding rule tied to a concrete timeline.

Voluntary frameworks with no enforcement mechanism are recorded as Framework entries rather than excluded.

Revision

The dataset is checked against primary regulatory texts and supervisory publications, and revised on activation events, material consultations, statutory amendments and confirmed corrections.

The revision date at the top of this page reflects the most recent edit, and each material change is recorded in the corrections log.

Independence

No jurisdiction, government, regulator, law firm or operator pays for placement, ranking or preferred treatment in this tracker. The dataset is published by Future Proof Intelligence under a CC BY 4.0 licence.

Corrections and updates can be sent through the contact page and are checked against primary sources before they are adopted.

Sources

All data comes from primary regulatory texts, such as official gazettes, parliamentary records and supervisory authority websites, or from official supervisory guidance. Secondary commentary is not used as a source.

Where data is unavailable or uncertain, a field reads "Not yet defined" or "TBD" rather than an estimate.

Penalty exposure

Penalty figures are the stated maximum in the primary instrument at the time of publication. They are not likely enforcement outcomes, typical settlements or legal advice.

Actual exposure depends on the facts, the jurisdiction and the regulator's discretion.

Extraterritorial reach

"Yes" means the primary instrument explicitly asserts jurisdiction over entities outside the territory whose systems affect people inside it. "Partial" means reach applies under specific conditions only.

"No" means the instrument does not assert extraterritorial application on its face, though conflict of laws analysis may still apply.

The EU AI Act timeline, date by date     The jurisdictions in depth