The most recent entries: 10 AI policy changes recorded across 8 jurisdictions, of which 2 instruments entered into force, 3 concerned binding law, 3 were rated urgent or high impact. Each is dated, links to its official source, and shows what it means in practice.
H.R. 10044 would tax large AI companies on the value of the tokens their models process or on AI-service revenue, whichever is higher, with a rate that rises automatically as unemployment climbs, and would create a Work Protection Administration in the Department of Labor to fund jobs. Introduced by Representatives Sara Jacobs, Greg Casar and Valerie Foushee. Recorded from secondary reporting; confirm on Congress.gov.
Practical impact: No obligations; a proposal to watch in the economic-transition tracker.
The general application date in Article 113 arrived. Chatbot disclosure, synthetic-content marking and deepfake labelling under Article 50 now apply; the high-risk obligations do not, having been deferred by Regulation (EU) 2026/1744. Recorded from secondary reporting; confirm against the Official Journal text.
Practical impact: Label AI interactions and synthetic content now; high-risk conformity work runs to the deferred dates.
Regulation (EU) 2026/1744, published in the Official Journal on 24 July 2026, entered into force on 27 July 2026. It amends Article 113 of the AI Act so that the Chapter III high-risk obligations apply from 2 December 2027 for stand-alone Annex III systems and from 2 August 2028 for AI embedded in Annex I products, adds two Article 5 prohibitions applying from 2 December 2026, and leaves the 2 August 2026 general application and Article 50 transparency dates in place. Recorded from concordant secondary reporting; confirm against the Official Journal text.
Practical impact: Re-plan high-risk conformity work to the December 2027 and August 2028 dates; transparency duties under Article 50 still applied from 2 August 2026.
Milestone recorded for Artificial Intelligence Basic Plan (Japan): Second AI Basic Plan adopted by Cabinet decision. Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for Artificial Intelligence (Malaysia): Ministry of Digital begins engagement and public consultation on the AI Governance Bill. Recorded from the official source listed; details to be expanded by a reviewer.
S. 4825 would levy a one-time tax, payable in equity, on AI companies with more than USD 200 million in annual sales, placing the shares in a public fund managed by an independent commission and paying dividends to residents. Introduced and referred; no committee action recorded. Recorded from secondary reporting and the sponsor's press release; confirm on Congress.gov.
Practical impact: No obligations; a proposal to watch in the economic-transition tracker.
Milestone recorded for Draft South Africa National Artificial Intelligence (South Africa): Withdrawal gazetted (Notice 3978 of 2026, Gazette No. 54840). Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for National Artificial Intelligence Strategy 2.0 (Singapore): Update to the National AI Strategy released with 10 refreshed priorities. Recorded from the official source listed; details to be expanded by a reviewer.
The Governor signed SB 26-189, which repeals and re-enacts the Colorado AI Act (SB 24-205) before its delayed 30 June 2026 effective date. The replacement law narrows the scope to automated decision-making technology that materially influences consequential decisions, drops the duty of care, impact assessments, the risk-management programme and Attorney General notification, and adds advance notice, adverse-decision disclosure, meaningful human review, three-year record keeping and developer documentation duties. Recorded from secondary reporting; confirm against the enrolled bill.
Practical impact: Stop building to the SB 24-205 impact-assessment and risk-programme requirements; plan notice, adverse-decision disclosure, human review, record keeping and developer documentation for consequential decisions from 1 January 2027.
Milestone recorded for Rancangan Peraturan Presiden tentang Peta Jalan Kecerdasan Artifisial Nasional 2026-2029 (Indonesia): Government completes inter-ministerial finalisation of the roadmap and ethics draft regulations. Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for Draft South Africa National Artificial Intelligence (South Africa): Minister Malatsi announces withdrawal of the draft policy. Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for Ghana National Artificial Intelligence Strategy (Ghana): Official launch in Accra; Minister's opening remarks published. Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for Plan Nacional de Inteligencia Artificial (Mexico): Plan Nacional de IA presented by ATDT (April 2026). Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for AI Guidelines for Business (Japan): Ver1.2 published. Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for Law on Artificial Intelligence (Vietnam): Law on Artificial Intelligence enters into force. Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for Rancangan Peraturan Presiden tentang Peta Jalan Kecerdasan Artifisial Nasional 2026-2029 (Indonesia): Inter-ministerial plenary discusses draft Perpres on National AI Roadmap 2026-2029. Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for National Artificial Intelligence Policy 2026-2030 (Bangladesh): Draft V2.0 published; consultation closed 8 February 2026. Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for National Artificial Intelligence Policy 2026-2030 (Bangladesh): Draft V1.1 released for public consultation. Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for Framework Act on the Development of Artificial Intelligence and Establishment of a Foundation for Trust (South Korea): Act enters into force. Recorded from the official source listed; details to be expanded by a reviewer.
The Working for Workers Four Act provisions took effect requiring employers with 25 or more employees to state in public job postings whether AI is used to screen, assess or select applicants.
Practical impact: Add an AI-use statement to every public job posting for Ontario roles.
Public Act 103-0804 became effective: discriminatory AI use in employment decisions is a civil-rights violation and employers must notify employees of AI use.
Practical impact: Employers must have notices and bias testing in place for AI tools affecting Illinois workers.
Milestone recorded for National AI Strategy (Egypt): NCAI mandate expanded to AI, Quantum Computing and Emerging Technologies (January 2026). Recorded from the official source listed; details to be expanded by a reviewer.
Most provisions of the Transparency in Frontier Artificial Intelligence Act took effect, including the duty to publish a frontier AI framework and report critical safety incidents.
Practical impact: Large frontier developers should now have a published framework and incident-reporting process.
Milestone recorded for Artificial Intelligence Basic Plan (Japan): First AI Basic Plan adopted by Cabinet decision. Recorded from the official source listed; details to be expanded by a reviewer.
The Governor signed the Responsible AI Safety and Education Act with agreed amendments aligning it to California SB 53, requiring large frontier developers to publish safety protocols and report critical incidents, reportedly effective 1 January 2027.
Practical impact: Frontier-model developers serving New York should align safety documentation with California SB 53 to satisfy both laws; verify the enrolled chapter text.
Milestone recorded for Law on Artificial Intelligence (Vietnam): Law No. 134/2025/QH15 passed by the National Assembly. Recorded from the official source listed; details to be expanded by a reviewer.
Milestone recorded for National Strategy on Research, Development and Application of Artificial Intelligence to 2030 (Vietnam): National Assembly passes the Law on Artificial Intelligence, creating a binding framework. Recorded from the official source listed; details to be expanded by a reviewer.
The National AI Plan set out the government's AI agenda and indicated that AI would be governed by strengthening existing laws rather than a standalone AI Act. The status of the 2024 mandatory-guardrails proposal must be confirmed.
Practical impact: Australian organisations should not expect a cross-economy AI Act in the near term; existing-law compliance and the voluntary standard remain the baseline.
The Commission proposed a "Digital Omnibus" package that would, among other changes, link the application of certain high-risk obligations to the availability of standards and support tools, with later backstop dates. This is a proposal and requires adoption by the European Parliament and Council.
Practical impact: High-risk timelines may move; organisations should keep planning against Article 113 dates until an amendment is adopted.
The most recent change on record occurred on 6 August 2026. The page is generated from the records, so it changes the moment a new entry is logged; the "Last updated" time above is the last edit to any record shown.
Where do these updates come from?
Every entry is written from an official source — a journal, a regulator, a legislature — and links to it. Nothing here is a press summary of a press summary. Entries record what changed, the status after the change and what it means in practice.
How is "top story" decided?
By a published rule, not an editor's mood: impact level, whether the instrument is binding, whether it entered into force, whether a named reviewer confirmed the record, and how recent it is, added up and shown as a score out of 100.
Can I follow updates for one country?
Yes. Every jurisdiction with recorded changes has its own updates page and RSS feed, and the weekly digest can be limited to the jurisdictions you choose.
Informational only, not legal advice. Verify every claim against the linked official sources and consult qualified counsel before acting.
Frequently asked questions
When was this page last updated?
The most recent change on record occurred on 6 August 2026. The page is generated from the records, so it changes the moment a new entry is logged; the "Last updated" time above is the last edit to any record shown.
Where do these updates come from?
Every entry is written from an official source — a journal, a regulator, a legislature — and links to it. Nothing here is a press summary of a press summary. Entries record what changed, the status after the change and what it means in practice.
How is "top story" decided?
By a published rule, not an editor's mood: impact level, whether the instrument is binding, whether it entered into force, whether a named reviewer confirmed the record, and how recent it is, added up and shown as a score out of 100.
Can I follow updates for one country?
Yes. Every jurisdiction with recorded changes has its own updates page and RSS feed, and the weekly digest can be limited to the jurisdictions you choose.
