AI & ML

China's Anthropomorphic-AI Rules Take Effect 15 July — and Doubao and Qwen Are Pulling Their Companion Agents

China's new framework for humanlike AI — the Interim Measures for the Administration of AI Anthropomorphic Interactive Services — takes effect on 15 July 2026, and rather than retrofit their products to comply, ByteDance's Doubao and Alibaba's Qwen are disabling their custom AI-agent and companion features. The rules do not ban companion AI, but they require AI disclosure, anti-addiction and exit controls, and protections for minors that persistent emotional-agent products were not built to operate unchanged.

6 Jul 2026 · 8 min read

AI & ML

A Bipartisan US Bill Wants One Federal Rulebook for Frontier AI — and the Fight Is Already About State Laws

Two House lawmakers have released a discussion draft of the Great American AI Act, a 269-page bipartisan attempt to set a single federal framework for governing frontier AI. It is not yet a bill, and its most contested feature — a three-year freeze on state AI laws — drew opposition within hours. For the region, the signal is what a US national standard would mean for the global rules AI developers plan around.

16 Jun 2026 · 8 min read

AI & ML

Brussels Disputes Apple's 'Due to DMA' Framing: EU Says the Siri AI Delay Is Apple's Choice, Not the Law's

A day after Apple said its new Siri AI would skip EU iPhones and iPads 'due to the DMA,' the European Commission pushed back: the delay is Apple's own business decision, the rules don't prohibit the launch, and Apple asked for an 18-month exemption rather than building a compliant solution. Apple says the Commission's demands would force unsafe access to user data. Both accounts are now on the record, and they don't agree.

11 Jun 2026 · 7 min read

AI & ML

After WWDC 2026, Apple's AI Is Fragmenting by Jurisdiction — Delayed in the EU, Unavailable in China

Apple's WWDC 2026 announcements made one thing clear for governance watchers: Siri AI and Apple Intelligence no longer ship as one global product. They are delayed in the EU under the DMA, unavailable in China pending regulatory requirements, and wrapped in a new age-based child-safety layer — regulation now decides what ships where.

9 Jun 2026 · 7 min read

AI & ML

Singapore Refreshes Its National AI Strategy: 10 Priorities, Four Missions, and a Bet Against a Two-Speed Economy

Singapore's 20 May update to its National AI Strategy refreshes the priorities across its 10 enablers, adds four sector 'missions', and sets out a programme to pull 10,000 enterprises past the experimentation stage — an explicit hedge against the 'K-shaped' AI economy the document itself warns about.

8 Jun 2026 · 7 min read

AI & ML

OpenAI Foundation commits US$250 million to workers displaced by AI

The OpenAI Foundation has pledged an initial US$250 million for research, grants and programmes to help workers and economies adjust to AI-driven job loss. It lands in a quarter when the tech industry cut close to 80,000 roles, roughly half of them tied to AI.

1 Jun 2026 · 3 min read

AI & ML

IMDA's Agentic AI Governance Framework v1.5 Names 60-Org Case Studies and Targets Multi-Agent Risk

Singapore's IMDA published Version 1.5 of its Model AI Governance Framework for Agentic AI on 20 May 2026 — a 51-page update grounded in feedback from more than 60 organisations including AWS, DBS, Google, and GovTech. The revision adds a six-category multi-agent risk taxonomy, a three-tier technical control structure, and concrete automation bias monitoring requirements that go beyond procedural checkbox compliance.

1 Jun 2026 · 5 min read

AI & ML

UAE Government 4.0: 80,000 Officials to Train in Agentic AI as Cabinet Greenlights Drafting of Federal Healthcare AI Law

The UAE Cabinet has approved the world's largest government agentic AI training programme — 80,000 federal employees across five tiers — and authorised drafting of a Federal Healthcare AI Law. Here is what the Cabinet actually decided, and why the healthcare law signal matters more than the headline number.

1 Jun 2026 · 6 min read

AI & ML

EU AI Act Omnibus: Annex III High-Risk Deadline Extends 16 Months, Annex I Gets 12, Nudifier Apps Banned from December 2026

A provisional political agreement reached on 7 May 2026 pushes Annex III standalone high-risk AI compliance to December 2027 — 16 months later — while Annex I product-embedded AI gets a 12-month reprieve to August 2028. A new prohibition on AI-generated non-consensual intimate imagery takes effect December 2026, regardless of the extensions.

1 Jun 2026 · 5 min read

AI & ML

Illinois Passes First US Law Requiring Independent Audits of Frontier AI — a Narrower Claim Than It Sounds

Senate Bill 315 cleared the Illinois House 110-0 on 27 May 2026, making Illinois the first US state to require annual independent third-party safety audits of large frontier AI developers — a specific requirement with no precedent in American law, even as two other states have already enacted their own frontier model standards.

1 Jun 2026 · 5 min read