1. What the AI policies of the three countries are (the short version)

• United States: Executive Order 14110 of 2023 was rescinded in January 2025. Since then, activity has run through executive orders (14179 in January 2025, 14365 in December, 14409 in June 2026, 14434 in September and others), state laws (California, Texas, Colorado, New York) and export controls Sourced [Ref. 1-12].
• China: Since 2022, departmental rules (algorithmic recommendation, deep synthesis, generative AI, labeling, anthropomorphic interaction) have come into force one after another, and the amended Cybersecurity Law took effect on 1 January 2026 Sourced [Ref. 32-39].
• Japan: The AI Promotion Act was promulgated on 4 June 2025 and fully in force from 1 September. The AI Strategy Headquarters sets a basic plan (23 December 2025; Phase II on 14 July 2026). For businesses there are the AI Guidelines for Business (version 1.2, 31 March 2026) Sourced [Ref. 40-48].

2. Reading types and stages of documents

The figure below sorts the documents of the three countries into "statute", "orders and rules", "guidance and plans" and "states and localities". It does not compare documents of different stages for strength or superiority.

The arrangement of stages is our reading. See each chapter and the original text for the binding force of each document.

3. United States: the flow of executive orders

In the United States, much of the federal activity on AI is set out in executive orders. We line up, in time order, the main executive orders whose signing dates could be confirmed in the Federal Register Sourced [Ref. 1-12].

A summary of the text in the Federal Register or on the White House site (Sourced). We did not check whether the deadlines set by the orders were met.

4. United States: a change of terminology — Executive Order 14434 and "Super Intelligence"

Executive Order 14434, "Inaugurating the Era of Super Intelligence" (29 September 2026), made it policy for the executive branch to use "Super Intelligence" and "SI" in place of "Artificial Intelligence" and "AI" in official documents. The main points of the text are as follows Sourced [Ref. 11, 12].

• Policy (section 1): to the maximum extent permitted by law, the executive branch is to use "Super Intelligence" and "SI" in place of "Artificial Intelligence" and "AI" and will not acknowledge the usage of the latter in any applicable setting.
• Implementation (section 2): agencies are to use the new terms in official correspondence, public communications, websites, reports, policy documents and other non-statutory documents. Nothing requires the alteration of previously issued regulations, Presidential actions, contracts, grants or other historical documents.
• Definition (section 3): "Super Intelligence" and "SI" mean the technologies and systems encompassed by the term "artificial intelligence" as defined in 15 U.S.C. 9401(3). Within 60 days, the Assistant to the President for Science and Technology is to submit proposed legislative language to establish a Federal definition, including an assessment of whether it should modify, expand or supersede the existing statutory definition.
• General provisions (section 4): it does not impair agencies' legal authority and creates no rights enforceable by private parties.

As of the check date, the new term appears on some US government sites. For example, the White House list page carries the order as "Inaugurating The Era Of Super Intelligence", and the top menu of NIST's site has a "Super intelligence" entry [Ref. 12, 30]. The body of the Bureau of Industry and Security (BIS) press release of 13 May 2025 also uses "SI" in its text as of the check date [Ref. 21]. We did not check whether the body of earlier announcements was revised after publication or was worded that way originally.

5. United States: federal law, state laws and Congress

Federal law

A federal statute specific to AI that we could confirm is the TAKE IT DOWN Act (Public Law 119-12, 19 May 2025). It criminalizes the intentional disclosure of nonconsensual intimate visual depictions (including "digital forgeries" made with software, machine learning or artificial intelligence), and requires "covered platforms", within one year of enactment, to set up a notice-and-removal process and to remove a depiction within 48 hours of a valid request, making reasonable efforts to remove known identical copies. The Federal Trade Commission (FTC) enforces it Sourced [Ref. 14]. On the other hand, we could not confirm the enactment of a comprehensive federal AI statute, or of a federal statute that generally preempts state AI laws. On 20 March 2026 the White House published legislative recommendations for Congress. They cover protecting children, communities and energy, intellectual property, speech, innovation, the workforce, and a federal framework with preemption of state law; they say that states should not be permitted to regulate AI development, and state the administration's belief that training AI models on copyrighted material does not violate copyright law while leaving the question to the courts Sourced [Ref. 13]. The recommendations are a request to Congress and are not legally binding.

State laws

EO 14365 made the relation to state law a policy issue, but a federal order does not by itself void state law; preemption requires a federal statute or a valid federal regulation (our reading). The main state laws we could confirm are as follows.

A summary of the official pages and texts of each legislature (Sourced). Check each legislature's page for the full text and the latest amendments.

By a memorandum of 9 January 2026, the Attorney General set up the AI Litigation Task Force, which is to challenge state AI laws on grounds that they "unconstitutionally regulate interstate commerce", are preempted by existing Federal regulations or are otherwise unlawful, and to consult White House officials [Ref. 9]. We make no guess about the outcomes of individual lawsuits over state laws, which are pending proceedings.

6. United States: export controls

Export controls on AI chips are set by the Bureau of Industry and Security (BIS) of the Department of Commerce in the Export Administration Regulations (EAR). We list the main developments confirmed in the Federal Register in time order Sourced.

A summary of statements in the Federal Register or official BIS documents (Sourced). Whether a licence is needed depends on the classification of the item (ECCN), the destination, and the end use and end user; check the original text and professional advice for any particular case.

7. United States: the name of NIST's AI evaluation unit

The name of NIST's AI evaluation unit has changed over time. The figure below follows NIST's public pages. The titles of documents such as the NIST AI RMF and NIST AI 600-1 have not changed (and are kept as originally titled here).

• August 2024: NIST announced that the U.S. AI Safety Institute had signed agreements on research, testing and evaluation with Anthropic and OpenAI [Ref. 27].
• June 2025: NIST's document of June 2025 says that, under the direction of the President, the Secretary of Commerce announced plans to reform the U.S. AI Safety Institute into the "Center for AI Standards and Innovation (CAISI)", which would serve as industry's primary point of contact within the U.S. Government for testing and collaborative research on commercial AI systems and operate within NIST [Ref. 26].
• As of the check date: NIST's site (nist.gov/caissi) names it the "Center for Advancing Innovation and Standards for Super Intelligence (CAISSI)". Its listed roles include being industry's primary point of contact, unclassified evaluations of capabilities that may pose national security risks (cybersecurity, biosecurity, chemical weapons) under voluntary agreements, evaluations and assessments of US and adversary systems, and engagement with international standards [Ref. 29]. NIST's hub page for the field (nist.gov/super-intelligence) also uses the new term [Ref. 30].
• On 17 February 2026, NIST announced the "AI Agent Standards Initiative" under the CAISI name (three pillars: industry-led agent standards, open-source protocols, and research on agent security and identity) [Ref. 28].

We could not find an announcement of the change of name, so documents under either name may be encountered. This article gives the name in use at the date of a document (CAISI) alongside the name on the current page (CAISSI). The role of evaluation bodies is covered in our explainer on AI safety evaluation and risk management.

8. China: generative AI, labeling and the Cybersecurity Law

China's AI regulation has advanced by layering departmental rules issued jointly by the Cyberspace Administration of China (CAC) and other bodies. We organize the main AI-related documents in order of entry into force. The originals were checked on the CAC's official site Sourced [Ref. 32-39].

A summary of the originals on the CAC site (Sourced). Check the original for the full text and the latest amendments.

The technical details of labeling are in the mandatory national standard GB 45438-2025, "Cybersecurity technology — Labeling method for content generated by artificial intelligence" (issued by the State Administration for Market Regulation and the Standardization Administration on 28 February 2025; in force 1 September 2025). The text of the standard (checked in a copy posted on a law firm's website) specifies, for example, that the text height of an explicit label must be at least 5% of the shortest side of the picture, that a video label must last at least 2 seconds at normal playback speed, and that the implicit metadata label is recorded in an extension field containing "AIGC" with the label type (1 = generated, 2 = possibly, 3 = suspected), the provider's name or code, a content number, the disseminator's name or code and a dissemination number. Implicit labels embedded in the content itself, such as digital watermarks, are not required by this standard (Annex A) Sourced [Ref. 39]. We could not confirm an official publication page for the standard.

The State Council issued its opinions on the "AI Plus" initiative (Guofa [2025] No. 11), dated 21 August 2025 and published on 26 August. They set targets: by 2027, penetration of new-generation intelligent terminals, agents and similar applications above 70%; by 2030, above 90%; and by 2035, entry into a new stage of an intelligent economy and society. They list actions in science and technology, industry, consumption, people's wellbeing, governance and international cooperation Not yet confirmed [Ref. 38]. These are policy targets, not results.

9. Japan: the AI Promotion Act, the basic plan and guidelines

The centre of Japan's AI policy is the "Act on the Promotion of Research and Development, and Utilization of Artificial Intelligence-Related Technologies" (Act No. 53 of 2025; the AI Promotion Act, called the AI Act in Japan). According to the e-Gov law data, it was promulgated on 4 June 2025; the parts other than Chapters III (the AI Basic Plan) and IV (the AI Strategy Headquarters) and a few supplementary provisions came into force on promulgation, and the rest on 1 September 2025 (the current version has an enforcement date of 1 September 2025) Sourced [Ref. 40, 41].

A summary of the text (e-Gov) (Sourced). The Act's chapters are General Provisions, Basic Measures, the AI Basic Plan, and the AI Strategy Headquarters; it has no chapter on penalties. It is a framework for promotion and appropriateness, not direct regulation of businesses (our reading).

The basic plan and the guidelines

• AI Basic Plan: the first plan was adopted by the Cabinet on 23 December 2025, and Phase II was adopted on 14 July 2026 (English provisional title: "Japan's Second Artificial Intelligence Basic Plan - Japan AI Transformation; A Stronger Nation. Greater Prosperity.") [Ref. 42-44]. Phase II, noting the rapid growth of agentic AI, organizes measures in four sections: Accelerating AI Deployment, Strengthening AI Development Capabilities, Leading AI Governance, and Building an AI-powered Society (official provisional English titles). Under AI governance it lists "responsible agile governance", a radical strengthening of the functions of the AI Safety Institute (AISI), engagement with the Hiroshima AI Process and international standards, support for technologies that identify AI-generated content (such as digital watermarks), and active review of institutions including the AI Act [Ref. 43]. The plan's measures and targets are Not yet confirmed and are not results.
• Guideline for Ensuring the Appropriateness of Research & Development and Utilization of Artificial Intelligence-Related Technology (provisional English title): decided by the Artificial Intelligence Strategic Headquarters (the AI Strategy Headquarters) on 19 December 2025 under Article 13 of the AI Act. They encourage voluntary and proactive efforts by all actors, state that they do not set "a single definition or absolute level" of appropriateness, and present elements such as human-centricity, fairness, safety, transparency, accountability, security and privacy. For research institutions and businesses utilizing AI, they list ensuring appropriateness through AI governance, ensuring transparency, ensuring sufficient safety, ensuring business continuity, and consideration of the importance of data [Ref. 45].
• AI Guidelines for Business (version 1.2): non-binding soft law published by MIC and METI on 31 March 2026 (see our explainer on deploying AI in enterprises) [Ref. 47, 48].

The appropriateness guidelines refer to the "Hiroshima AI Process", launched at the G7 Hiroshima Summit in 2023. They state that the Hiroshima AI Process Friends Group, launched in May 2024, had 60 countries and regions participating (as of December 2025), and that 24 organizations had submitted responses to the "reporting framework" in which developers voluntarily report on their adherence to the international code of conduct (same date) [Ref. 45]. The AI Strategy Headquarters is headed by the Prime Minister, and all Ministers of State are members [Ref. 46].

10. Where this touches materials and manufacturing

The policies of the three countries touch not only companies that build AI but also materials and manufacturing companies that use it (our reading). We organize what can be read from the texts and plans.

11. Limits, and what we could not confirm

(1) Limits of comparison

The documents of the three countries differ in legal character and in how they are changed. This article only lines up stages and dates and does not evaluate strength or superiority.

(2) Implementation and pending proceedings

We could not confirm whether the deadlines in executive orders were met (publication of Commerce's evaluation of state laws, the BEAD notice, deliverables within the deadlines of EO 14409 and 14434). We make no guess about the outcomes of lawsuits over state laws or of bills in Congress on preemption (including drafts).

(3) Originals we could not check

For China's amended Cybersecurity Law, we confirmed the content in the CAC's commentary and did not check the text of the amending decision. For GB 45438-2025, we checked a copy of the standard's text on a law firm's site and could not confirm an official publication page. The Department of Commerce's June 2025 statement (by the Secretary) could not be reached on its site, and we confirmed the plan in NIST's document instead.

(4) The history of names

We could not find an announcement giving the date or reasons for NIST's unit becoming CAISSI. How far EO 14434 changes the term for US "AI" may change after the definition proposal under section 3(b), so we keep to what the current text says.

(5) Countries and fields not covered

The EU is covered in our explainer on the EU AI Act. We do not cover the policies of the United Kingdom, Korea, India and others, or AI policy in the military field.

12. How to read evaluations and standards

Policy documents contain few numerical benchmarks. Instead there are thresholds that decide what is regulated (compute, performance), evaluations carried out by businesses or governments, and standards and guidance used to show conformity. The table below lines them up as far as they could be confirmed; where nothing could be confirmed the cell says so.

Checked against laws, the Federal Register, official government pages and the text of a standard (Sourced; checked 15 October 2026). The arrangement of this table is Our calculation (a reordering by this article), and the reading-note column is commentary.

Notes on reading

• A threshold means different things in each system. California's 10²⁶ FLOP, the EU's 10²⁵ FLOP presumption, the TPP in US export controls and China's security-assessment scope are criteria for different purposes, and are not lined up to compare "strictness".
• Check whether a document is mandatory or voluntary. China's mandatory national standard is mandatory, the NIST AI RMF is voluntary, Japan's AI Guidelines for Business are non-binding, and a US executive order is an instruction to the executive branch.
• Names and dates change. The NIST unit's name, AI-related terms (Executive Order 14434) and the version of the basic plan have all changed. When citing, keep the document's title as originally given and add the check date and version.

13. Glossary

14. References

1. US President / Federal Register Executive Order 14148 "Initial Rescissions of Harmful Executive Orders and Actions" (signed 20 January 2025; Federal Register 28 January 2025; the list of rescinded orders includes EO 14110) — www.federalregister.gov/documents/2025/01/28/2025-01901/initial-rescissions-of-harmful-executive-orders-and-actions
2. US President / Federal Register Executive Order 14179 "Removing Barriers to American Leadership in Artificial Intelligence" (signed 23 January 2025; published 31 January 2025) — www.federalregister.gov/documents/2025/01/31/2025-02172/removing-barriers-to-american-leadership-in-artificial-intelligence
3. The White House "America's AI Action Plan" (July 2025) — www.whitehouse.gov/wp-content/uploads/2025/07/Americas-AI-Action-Plan.pdf
4. US President / Federal Register Executive Order 14318 "Accelerating Federal Permitting of Data Center Infrastructure" (signed 23 July 2025) — www.federalregister.gov/documents/2025/07/28/2025-14212/accelerating-federal-permitting-of-data-center-infrastructure
5. US President / Federal Register Executive Order 14319 "Preventing Woke AI in the Federal Government" (signed 23 July 2025) — www.federalregister.gov/documents/2025/07/28/2025-14217/preventing-woke-ai-in-the-federal-government
6. US President / Federal Register Executive Order 14320 "Promoting the Export of the American AI Technology Stack" (signed 23 July 2025) — www.federalregister.gov/documents/2025/07/28/2025-14218/promoting-the-export-of-the-american-ai-technology-stack
7. US President / Federal Register Executive Order 14363 "Launching the Genesis Mission" (signed 24 November 2025; published 28 November 2025) — www.federalregister.gov/documents/2025/11/28/2025-21665/launching-the-genesis-mission
8. US President / Federal Register Executive Order 14365 "Ensuring a National Policy Framework for Artificial Intelligence" (signed 11 December 2025; published 16 December 2025) — www.federalregister.gov/documents/2025/12/16/2025-23092/ensuring-a-national-policy-framework-for-artificial-intelligence
9. US Department of Justice Attorney General memorandum "Artificial Intelligence Litigation Task Force" (9 January 2026) — justice.gov/ag/media/1422986/dl
10. The White House Executive Order 14409 "Promoting Advanced Artificial Intelligence Innovation and Security" (signed 2 June 2026; Federal Register document 2026-11415) — www.whitehouse.gov/presidential-actions/2026/06/promoting-advanced-artificial-intelligence-innovation-and-security/
11. US President / Federal Register Executive Order 14434 "Inaugurating the Era of Super Intelligence" (signed 29 September 2026; published 2 October 2026) — www.federalregister.gov/documents/2026/10/02/2026-20321/inaugurating-the-era-of-super-intelligence
12. The White House "Inaugurating The Era Of Super Intelligence" (page of Executive Order 14434, 29 September 2026; also listed on the executive orders page) — www.whitehouse.gov/presidential-actions/2026/09/inaugurating-the-era-of-super-intelligence/
13. The White House "National Policy Framework for Artificial Intelligence: Legislative Recommendations" (20 March 2026) — www.whitehouse.gov/wp-content/uploads/2026/03/03.20.26-National-Policy-Framework-for-Artificial-Intelligence-Legislative-Recommendations.pdf
14. US Congress / Government Publishing Office Public Law 119-12, "TAKE IT DOWN Act" (S. 146, 19 May 2025) — www.govinfo.gov/content/pkg/PLAW-119publ12/html/PLAW-119publ12.htm
15. California Legislature SB-53 "Artificial intelligence models: large developers" (bill status page; approved by the Governor on 29 September 2025, Chapter 138, Statutes of 2025) — leginfo.legislature.ca.gov/faces/billStatusClient.xhtml?bill_id=202520260SB53
16. California Legislature SB-53, chaptered bill text — leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=202520260SB53
17. Texas Legislature 89(R) HB 149 "Texas Responsible Artificial Intelligence Governance Act" (bill history page; the enrolled text is at capitol.texas.gov/tlodocs/89R/billtext/html/HB00149F.htm) — capitol.texas.gov/BillLookup/History.aspx?LegSess=89R&Bill=HB149
18. Colorado General Assembly SB26-189 "Automated Decision-Making Technology" (bill page; Signed Act, 14 May 2026) — leg.colorado.gov/bills/sb26-189
19. New York State Legislature S6953B, relating to the training and use of artificial intelligence frontier models (signed 19 December 2025, Chapter 699) — www.nysenate.gov/legislation/bills/2025/S6953
20. New York State Legislature S8828 (chapter amendment; signed 27 March 2026, Chapter 96) — www.nysenate.gov/legislation/bills/2025/S8828
21. US Department of Commerce, BIS "Department of Commerce Announces Rescission of Biden-Era Artificial Intelligence Diffusion Rule, Strengthens Chip-Related Export Controls" (press release dated 13 May 2025; the body as checked uses the term "SI") — www.bis.gov/press-release/department-commerce-announces-rescission-biden-era-artificial-intelligence-diffusion-rule-strengthens
22. BIS / Federal Register "Framework for Artificial Intelligence Diffusion", 90 FR 4544 (15 January 2025) — www.federalregister.gov/documents/2025/01/15/2025-00636/framework-for-artificial-intelligence-diffusion
23. BIS / Federal Register "Revision to License Review Policy for Advanced Computing Commodities", 91 FR 1684 (15 January 2026) — www.federalregister.gov/documents/2026/01/15/2026-00789/revision-to-license-review-policy-for-advanced-computing-commodities
24. BIS "Guidance Regarding Enforcement of License Requirements for Advanced Computing Items for Entities Headquartered in Country Group D:5 and Macau" (31 May 2026) — www.bis.gov/media/documents/bis-guidance-may-31-2026.pdf
25. BIS / Federal Register "Enhanced Favorable Treatment for the United Arab Emirates Under the Export Administration Regulations", 91 FR 43034 (14 July 2026; effective 10 July) — www.federalregister.gov/documents/2026/07/14/2026-14132/enhanced-favorable-treatment-for-the-united-arab-emirates-under-the-export-administration
26. NIST "NIST Announcements" (June 2025; includes the plan to reform the U.S. AI Safety Institute into CAISI) — www.nist.gov/document/nist-announcements-june-2025
27. NIST "U.S. AI Safety Institute Signs Agreements Regarding AI Safety Research, Testing and Evaluation With Anthropic and OpenAI" (August 2024) — www.nist.gov/news-events/news/2024/08/us-ai-safety-institute-signs-agreements-regarding-ai-safety-research
28. NIST "Announcing the 'AI Agent Standards Initiative' for Interoperable and Secure Innovation" (17 February 2026; announced under the CAISI name) — www.nist.gov/news-events/news/2026/02/announcing-ai-agent-standards-initiative-interoperable-and-secure
29. NIST "Center for Advancing Innovation and Standards for Super Intelligence (CAISSI)" (current page as checked 15 October 2026; its news list carries evaluation articles under the CAISI name) — www.nist.gov/caissi
30. NIST NIST hub page on AI ("Super intelligence"), as checked 15 October 2026 — www.nist.gov/super-intelligence
31. NIST "AI Risk Management Framework" (as checked 15 October 2026; states that AI RMF 1.0 is being revised) — www.nist.gov/itl/ai-risk-management-framework
32. Cyberspace Administration of China and others Interim Measures for the Administration of Generative Artificial Intelligence Services, Order No. 15 (signed 10 July 2023; in force 15 August 2023; in Chinese) — www.cac.gov.cn/2023-07/13/c_1690898327029107.htm
33. Cyberspace Administration of China and others Provisions on the Administration of Deep Synthesis in Internet Information Services, Order No. 12 (signed 25 November 2022; in force 10 January 2023; in Chinese) — www.cac.gov.cn/2022-12/11/c_1672221949354811.htm
34. Cyberspace Administration of China and others Provisions on the Administration of Algorithmic Recommendations in Internet Information Services, Order No. 9 (signed 31 December 2021; in force 1 March 2022; in Chinese) — www.cac.gov.cn/2022-01/04/c_1642894606364259.htm
35. Cyberspace Administration of China and others Measures for Labeling AI-Generated Synthetic Content, CAC Notice [2025] No. 2 (dated 7 March 2025; in force 1 September 2025; in Chinese) — www.cac.gov.cn/2025-03/14/c_1743654684782215.htm
36. Cyberspace Administration of China and others Interim Measures for the Administration of Anthropomorphic AI Interaction Services, Order No. 21 (issued 10 April 2026; in force 15 July 2026; in Chinese) — www.cac.gov.cn/2026-04/10/c_1777558395078289.htm
37. Cyberspace Administration of China Expert commentary on the amendment of the Cybersecurity Law (2 January 2026; in Chinese; explains the decision adopted on 28 October 2025, in force 1 January 2026, and the new Article 20) — www.cac.gov.cn/2026-01/02/c_1769093523928606.htm
38. State Council of China State Council Opinions on Deepening the Implementation of the "AI Plus" Initiative, Guofa [2025] No. 11 (dated 21 August 2025; published 26 August; in Chinese) — www.gov.cn/zhengce/content/202508/content_7037861.htm
39. State Administration for Market Regulation and Standardization Administration of China GB 45438-2025, Cybersecurity technology - Labeling method for content generated by artificial intelligence (issued 28 February 2025; in force 1 September 2025). We used a copy of the standard's text posted on a law firm's website — sidley.com (PDF copy of GB 45438-2025)
40. Japan (e-Gov Law Search) Act on the Promotion of Research and Development, and Utilization of Artificial Intelligence-Related Technologies (Act No. 53 of 2025; promulgated 4 June 2025; current version in force from 1 September 2025; in Japanese) — laws.e-gov.go.jp/law/507AC0000000053
41. Cabinet Office, Japan Page on the AI Promotion Act (in Japanese) — www8.cao.go.jp/cstp/ai/ai_act/ai_act.html
42. Cabinet Office, Japan Page on the AI Basic Plan (Cabinet decision of 14 July 2026; earlier plan: Cabinet decision of 23 December 2025; in Japanese) — www8.cao.go.jp/cstp/ai/ai_plan/ai_plan.html
43. Cabinet Office, Japan Japan's Second Artificial Intelligence Basic Plan (Cabinet decision of 14 July 2026; in Japanese, with an official provisional English translation) — www8.cao.go.jp/cstp/ai/ai_plan/aiplan_20260714.pdf
44. Cabinet Office, Japan Artificial Intelligence Basic Plan (Cabinet decision of 23 December 2025; in Japanese) — www8.cao.go.jp/cstp/ai/ai_plan/aiplan_20251223.pdf
45. Artificial Intelligence Strategic Headquarters, Japan Guideline for Ensuring the Appropriateness of Research & Development and Utilization of Artificial Intelligence-Related Technology (decision of 19 December 2025; in Japanese, with an official provisional English translation) — www8.cao.go.jp/cstp/ai/ai_guideline/ai_gl_2025.pdf
46. Cabinet Office, Japan Page of the Artificial Intelligence Strategic Headquarters (membership; in Japanese) — www8.cao.go.jp/cstp/ai/ai_hq/ai_hq.html
47. MIC and METI, Japan AI Guidelines for Business, Version 1.2 (31 March 2026; in Japanese) — www.meti.go.jp/shingikai/mono_info_service/ai_shakai_jisso/pdf/20260331_1.pdf
48. METI, Japan Page of the Study Group on AI Guidelines for Business (list of publication dates for versions 1.0 to 1.2; in Japanese) — www.meti.go.jp/shingikai/mono_info_service/ai_shakai_jisso/index.html

15. Claim-to-source audit

Last updated: 15 October 2026 / Sources are limited to primary material (the Federal Register, the White House, official pages of the US Congress and state legislatures, official documents of BIS and NIST, the CAC, the State Council and the central government portal of China, the e-Gov law database, and the Cabinet Office, MIC and METI of Japan). The arrangement of document stages and the readings for materials and manufacturing are marked as commentary or our calculation to keep them apart from sourced facts. Whether the deadlines in executive orders were met, publication of Commerce's evaluation of state laws, the original text of the amended Cybersecurity Law, an announcement of NIST's change of unit name, and an official publication page for GB 45438-2025 could not be confirmed and are not described. This article is not legal advice and neither supports nor criticizes any country, administration or company. We make no guess about the outcome of pending proceedings or future legislation. Figures 2 to 4 are vector drawings; Figures 1 and 5 are frames for AI-generated images (not yet created). None of them depicts a real organization or document.