Free · Dated · Honestly incomplete
The AI disclosure map.
Which jurisdictions require disclosure or labeling of AI-generated content. Colored tiles are recorded — confirmed against primary or authoritative sources, with dates. Dashed tiles are on the watchlist — reported activity awaiting confirmation. Jurisdictions appear on this map only when the register first records or watches them; the rest of the world is not yet examined, which is different from empty.
Select a tile for its register entry. Solid tiles are confirmed entries; dashed tiles are watchlist items — reported, not yet confirmed.
Companion panel: United States, state by state →
Inclusion policy: a jurisdiction appears when the register first records or watches it — currently 4 recorded · 4 on watch · the rest not yet examined.
The record behind the map
| Jurisdiction | Instrument | Status | Key date | Source |
|---|---|---|---|---|
| European Union | AI Act Article 50 — interaction disclosure, machine-readable marking, biometric notice, deepfake & public-interest text labelling | Deadline ahead | 2 AUG 2026 (marking: 2 DEC 2026 for pre-existing systems) |
EUR-Lex |
| United States | State-level regimes: Texas TRAIGA and California AB 2013 in force (Jan 2026); Utah AI Policy Act in force (2024); California SB 942 watermarking operative 2 Aug 2026; Colorado AI Act (SB 24-205/SB 26-189) effective 1 Jan 2027, AG rulemaking pending. No federal disclosure statute recorded. | In force + deadline | IN FORCE · JAN 2026 SB 942: 2 AUG 2026 |
CA SB 942 |
| China | CAC Measures for Labeling AI-Generated Content + mandatory national standard GB 45438-2025 — explicit (visible) and implicit (metadata) labels; platform enforcement active | In force | 1 SEP 2025 | Translation |
| South Korea | AI Basic Act — Art. 31 transparency duties (AI-use notice, generative-output labeling, disclosure for outputs difficult to distinguish from reality) | In force | 22 JAN 2026 | LOC Global Legal Monitor |
| Canada Watchlist | Federal AI legislation reported in flux following the lapse of the previous framework bill. Not yet confirmed against primary sources. | Watching | — | Pending |
| United Kingdom Watchlist | Principles-based approach without a disclosure statute reported; consultation activity ongoing. Not yet confirmed. | Watching | — | Pending |
| Brazil Watchlist | A comprehensive AI bill reported advancing through the legislature. Not yet confirmed. | Watching | — | Pending |
| India Watchlist | Amendments to information-technology rules concerning labeling of synthetic media reported for 2026. Not yet confirmed. | Watching | — | Pending |
| All other jurisdictions | Not yet examined. Entries appear here only once recorded against primary or authoritative sources, or placed on the watchlist with a stated basis — each with the date first confirmed. | Not yet examined | — | — |
This map records disclosure and labeling regimes for AI-generated content specifically — not general AI regulation, high-risk frameworks, or sector rules. "Not yet examined" means exactly that — never "no law exists." Corrections and additions with primary sources: [email protected].