Ctrl+AI+Reg - 12 July 2026
Your shortcut to AI regulation, law and policy updates around the world.
AI Regulation Updates
In this issue:
Weekly Trend Analysis (for paid subscribers).
Updates from EU, UK, Italy, Spain, US, South Korea, Singapore, Indonesia, Thailand, China, Japan, Hong Kong, Australia, Nigeria.
See fuller updates in the ‘Global Updates’ tab in the Global AI Regulation Tracker (English version | Chinese version)
Previous update
Weekly Trend Analysis
Governments regulate AI more easily where the product touches their own citizens than where it crosses borders. Last week showed both sides of that gap.
The US Commerce Department received 78 applications under the American AI Exports Program, each proposing an end to end package of hardware, data, models and cybersecurity, with federal financing and expedited licensing for the winners. Days later, the Financial Times reported that OpenAI and Google had supplied AI services to Chinese organisations on a Pentagon list of entities with alleged military ties, via their Singapore units. Export controls assume goods that physically cross a border. Access to a hosted model does not. China faces the mirror problem, issuing a cybersecurity alert on Anthropic’s Claude Code that warned of a possible backdoor without technical detail or enforcement action.
Domestic harms drew faster action, and three unlike systems converged: