Library
The Global AI & Data Regulation Calendar
Next hard date after 13 August 2026: 2 December 2026. EU Article 50(2) marking grace ends, and two new Article 5 prohibitions apply. The table below is every live and upcoming deadline we currently track from the EU, UK, US and China pages. It will move. That is why the page exists.
Last reviewed
MikeFounder & developer
- Regulation calendar
- EU AI Act
- UK AI
- US & China
- Compliance
On this page
Who this is for
- Anyone who wants one bookmark instead of six trackers
- Operators planning a quarter who need the next statutory date, not a narrative
Note: Next deadline: 2 December 2026 (EU). Article 50(2) machine-readable marking for generative systems already on the market before 2 August 2026; new Article 5 prohibitions on non-consensual intimate imagery and AI-generated CSAM. Then 1 January 2027 (Colorado SB 26-189). Winter 2026 ICO ADM guidance is listed by the ICO without a day.
How to read the table
Sorted by date. Live means the duty already applies. Upcoming means it has not yet. Source pages: EU timeline, UK who enforces what, DUAA ADM, ICO code tracker, US state tracker, China rules.
This is not legal advice and not a complete world calendar. If a cell is uncertain, the source page says so.
Calendar
| Date | Status | Jurisdiction | What happens | Source page |
|---|---|---|---|---|
| 15 August 2023 | Live | China | Interim Measures for generative AI services in force | China |
| 1 August 2024 | Live | EU | AI Act (2024/1689) enters into force | EU timeline |
| 2 February 2025 | Live | EU | Article 5 original prohibitions; Article 4 AI literacy | EU timeline |
| 2 August 2025 | Live | EU | GPAI Chapter V (Arts 51–55) applies | GPAI |
| 1 September 2025 | Live | China | Identification Measures + GB 45438-2025 labelling | China |
| 5 February 2026 | Live | UK | DUAA s.80 / UK GDPR Arts 22A–22D (SI 2026/82) | DUAA ADM |
| 12 May 2026 | Live | UK | SI 2026/425: duty on ICO to prepare AI and ADM code (code itself not published) | ICO tracker |
| 14 May 2026 | Live (signed) | US (Colorado) | SB 26-189 signed; most duties apply 1 January 2027 | US tracker |
| 29 May 2026 | Live (closed) | UK | ICO ADM guidance consultation closed | ICO tracker |
| 27 July 2026 | Live | EU | Digital Omnibus on AI (2026/1744) in force | EU timeline |
| 2 August 2026 | Live | EU | Article 50 transparency; general application; AI Office GPAI enforcement | EU timeline |
| 2 August 2026 | Live | US (California) | SB 942 / BPC Chapter 25 operative (covered generative providers) | US tracker |
| Winter 2026 | Upcoming (no day) | UK | ICO lists final ADM guidance as due winter 2026. Not the statutory code. | ICO tracker |
| 2 December 2026 | Upcoming (next hard date) | EU | Art. 50(2) marking grace ends for pre-2 August systems; new Art. 5(ba) and (bb) | EU timeline |
| 1 January 2027 | Upcoming | US (Colorado) | SB 26-189 applies to consequential decisions; AG rules due | US tracker |
| 2 August 2027 | Upcoming | EU | GPAI models on the market before 2 August 2025 must comply; national sandboxes operational | EU timeline |
| 2 December 2027 | Upcoming | EU | Annex III / Art. 6(2) high-risk Chapter III duties | EU timeline |
| 2 August 2028 | Upcoming | EU | Annex I / Art. 6(1) high-risk Chapter III duties | EU timeline |
| 2 August 2030 | Upcoming | EU | High-risk systems intended for public authorities (Art. 111 as amended) | EU timeline |
| 31 December 2030 | Upcoming | EU | Annex X large-scale IT systems (original Art. 113; not amended in 2026/1744 as checked) | EU timeline |
Not on a date, still live to watch
| Item | Status as of 13 August 2026 | Source page |
|---|---|---|
| UK dedicated AI bill | None before Parliament | UK map |
| ICO statutory AI/ADM code (as opposed to guidance) | Duty live; draft not published; consultation date unclear | ICO tracker |
| EU-US Data Privacy Framework | In force; CJEU appeal pending (Latombe, C-703/25 P) | Transfers |
| Colorado xAI v. Weiser stay | Reported stay of original SB 24-205 enforcement; confirm on the docket | US tracker |
Questions
Why is 2 December 2026 “next” when ICO guidance is “winter 2026”?
Winter 2026 has no day on the ICO plan. 2 December 2026 is a date in Regulation (EU) 2026/1744. Hard dates sort above season labels.
Will you email me when a row moves?
No. Revisit the page. The last-reviewed chip and the changelog below are the signal. The argument for dating the library is in Why this site now has a library.
Where is the rest of the world?
Not on this table yet. If you operate there, do not infer silence means absence. This calendar is the six source pages, on purpose.
What changed
- 13 August 2026: First publication, synthesised from the six jurisdiction/tracker pages. Cadence: high-risk / fast-moving (30–90 days). Next review due 13 October 2026. Off-cycle when any source page's date moves.
Note: This is compliance consulting and training, not legal advice. We work alongside your legal counsel, translating regulation into operational reality rather than replacing them. For a structured pass, see AI compliance.
Was this helpful?
Also in the library
- What Counts as High-Risk Under Annex IIIAnnex III is a use-case list, not a model list. Most Lathestone clients who over-classify do it because the tool is 'AI'; most who under-classify miss internal HR and customer-decision systems. Article 6(3) exemptions exist. They are narrow, and profiling is not one of them.
- China's Generative AI Rules: Filing, Labelling, Content ControlPublic generative AI services in China need a CAC security assessment and algorithm filing before launch. Labelling of AI-generated content has been mandatory under the March 2025 Identification Measures and GB 45438-2025 since 1 September 2025. Selling into China is not the same as training or deploying a model there.
- The Data (Use and Access) Act 2025 and the New UK ADM RegimeSection 80 of the Data (Use and Access) Act 2025 replaced UK GDPR Article 22 with Articles 22A–22D from 5 February 2026. The old default was prohibition-plus-exceptions. The new default for most significant automated decisions is permission-plus-safeguards. Special-category decisions stay tighter.