Good morning. Here is your complete briefing on the day in artificial intelligence: what happened, the context behind it, and why it matters. 7 stories, two minutes, no noise. 01 | CHINA'S AI COMPANION BAN IS NOW LIVE, AND MILLIONS OF AGENTS ARE GONE China's Interim Measures for AI Anthropomorphic Interactive Services took effect today. ByteDance's Doubao, the country's most used AI app with 345 million monthly active users, pulled its agent features this morning, and Alibaba's Qwen shut down user created agents last week with configurations and chat histories permanently deleted. Doubao users keep read only access to their data until October 15, then it is gone. ByteDance is steering users to Maoxiang, a separate companion app, while Alibaba has offered no migration path at all. The rules target emotional companionship AI that remembers you and builds a relationship, while task focused assistants are exempt. Why it matters: The shutdowns previewed last week are now real. Hundreds of millions of users just watched their AI relationships get switched off by regulation overnight, the clearest demonstration yet of what hard AI rules feel like on the ground. 02 | NEW YORK BECOMES THE FIRST STATE TO FREEZE NEW AI DATA CENTERS Governor Kathy Hochul imposed a one year statewide moratorium on new hyperscale data centers of 50 megawatts or more, the first freeze of its kind in the nation. During the pause, the state will produce an environmental impact statement setting standards on grid impact, water use and air quality, and Hochul says future data centers will shoulder more of the infrastructure costs they create so utility bills stop climbing for residents. Critics warn billions in AI investment will simply move to friendlier states. Why it matters: The AI buildout just hit its first hard state level wall. If the New York model spreads, siting and power become as strategic as chips. 03 | HASSABIS CALLS FOR A US LED WATCHDOG WITH THE POWER TO PAUSE AI Google DeepMind CEO Demis Hassabis published a personal manifesto, A Framework for Frontier AI and the Dawning of a New Age, proposing an AI standards body modeled on FINRA, the industry funded regulator that polices Wall Street. Frontier labs would share models up to 30 days before release for safety testing on cyber, biological and deception capabilities, and the body could eventually gate access to the US market and coordinate an industry wide slowdown if dangers mount. The rules would apply to all frontier class models, foreign or domestic, open or closed.