Skip to main content

The Race

Who regulated what, and when

Ten years, three regulatory philosophies, one accelerating calendar. Brussels, Washington and Beijing on the same track from 2016 to 2026, so the order of events is a matter of looking rather than of argument.

Years covered
2016 to 2026
Events plotted
31
Lanes
3
Collisions
3
Firsts settled
4

Exhibit one

The race, jurisdiction by jurisdiction

The same events as the desktop track, listed in date order for each lane.

EU

European Union Β· 10 events

  1. 27 APR 2016

    GDPR adopted

    Art. 22's automated-decision right becomes the AI Act's legal ancestor.

  2. APR 2019

    Ethics Guidelines for Trustworthy AI

    The policy vocabulary is set: human agency, robustness, privacy, fairness.

  3. 21 APR 2021

    AI Act proposed

    First horizontal AI law attempt anywhere.

  4. 30 NOV 2022

    ChatGPT launches

    A GPAI chapter gets bolted onto the draft mid-negotiation.

  5. 8 DEC 2023

    Trilogue deal

    Facial-recognition bans and GPAI rules settled at the last minute.

  6. 1 AUG 2024

    AI Act enters into force

    AI Office already set up (Jan 2024).

  7. 2 FEB 2025

    First wave bites

    Prohibited practices banned EU-wide + AI-literacy duty.

  8. 2 AUG 2025

    GPAI + governance + penalties in force

    Code of Practice becomes the compliance route.

  9. 27 JUL 2026

    Digital Omnibus in force

    Reg. (EU) 2026/1744 defers high-risk duties to Dec 2027/Aug 2028.

  10. 2 AUG 2026 Β· NEXT UP

    Transparency + GPAI enforcement switch on

    Chatbots disclose, deepfakes labeled, Brussels can fine model providers.

US

United States Β· 10 events

  1. OCT 2016

    Obama OSTP report

    First federal AI strategy framing.

  2. 11 FEB 2019

    EO 13859

    First AI executive order; innovation-first.

  3. 14 MAY 2019

    San Francisco bans government facial recognition

    First major US city to ban it.

  4. 1 JAN 2020

    Illinois AIVIA

    First US AI-hiring law (video interviews).

  5. 26 JAN 2023

    NIST AI RMF 1.0

    The durable cross-administration standard.

  6. 30 OCT 2023

    Biden EO 14110

    High-water mark of federal AI safety (revoked 2025).

  7. 17 MAY 2024

    Colorado SB 24-205 signed

    First comprehensive state AI law (later: stayed, repealed).

  8. 20 to 23 JAN 2025

    The pivot

    EO 14110 revoked; EO 14179 signed; deregulation begins.

  9. JUL 2025

    The deregulation summer

    Senate 99 to 1 strips the state-law moratorium (1st); AI Action Plan released (23rd).

  10. JAN to MAY 2026

    State layer goes live; Colorado collapses

    TRAIGA + CA laws in force 1 Jan; xAI suit, DOJ intervention, stay, repeal Apr to May.

CN

China Β· 11 events

  1. 20 JUL 2017

    New Generation AI Development Plan

    Three-step roadmap to 2030 world leadership.

  2. JUN 2019

    Governance Principles

    First soft-law AI ethics statement.

  3. SEP to NOV 2021

    Governance stack

    Data Security Law + PIPL (Art. 24/26) enacted.

  4. 1 MAR 2022

    Algorithm Recommendation Provisions

    World's first operational algorithm registry opens.

  5. 10 JAN 2023

    Deep Synthesis Provisions

    First binding deepfake rules anywhere.

  6. 15 AUG 2023

    GenAI Interim Measures

    First national GenAI rulebook; pre-launch filing begins (61 services by year-end).

  7. 18 OCT 2023

    Global AI Governance Initiative

    Launched at the Belt & Road Forum.

  8. 1 JUN 2025

    Facial Recognition Measures

    Commercial FR restricted; 100k-faces filing duty.

  9. 26 JUL 2025

    WAIC counter-offer

    13-point Action Plan + proposed Global AI Cooperation Organization (Shanghai).

  10. 1 SEP 2025

    Labeling Measures + GB 45438-2025

    Dual explicit/implicit AI-content labels.

  11. 15 JUL 2026

    Anthropomorphic AI Measures in force

    Doubao/Qwen companion features shut down at the deadline.

Collisions

2023: China's GenAI filing regime opens while Washington drafts non-binding blueprints.

JUL 2025: US AI Action Plan (23rd) β†’ China's Global Action Plan (26th): the counter-offer.

2 AUG 2026: Brussels transparency duties + California watermarking: same deadline.

Exhibit two

The firsts

Four claims to being first, each with the instrument that settles it and the runner up that did not.

01

First binding algorithm rules

CNMAR 2022

Algorithm Recommendation Provisions

Runner up:EU AI Act proposed Apr 2021 but in force 2024; US: none comparable.

Open in the Explorer
02

First deepfake law

CNJAN 2023

Deep Synthesis Provisions

Runner up:US first federal content law only May 2025 (TAKE IT DOWN); EU Art. 50 duties 2 Aug 2026.

Open in the Explorer
03

First horizontal, comprehensive AI statute

EUAUG 2024

EU Artificial Intelligence Act

Runner up:CN: still unenacted; US: none federal; Colorado's never took effect.

Open in the Explorer
04

First AI-companion rulebook

CNJUL 2026

Anthropomorphic AI Measures

Runner up:US: CA SB 243 (Jan 2026) just behind; EU: none.

Open in the Explorer

China regulated first and fastest, vertically. The EU regulated broadest, horizontally. The United States regulated last at the federal level and loudest at the state level.

Now watch the same three regimes rule on real AI

The calendar tells you who moved. The simulator tells you what each regime would do to a system you put in front of it, and the explorer gives you the text behind every verdict.

Knowing the order of events is the start. Working inside these rules is the job.