← All Issues
Sunday, September 13, 2026

AI Daily

Your Automated Intelligence Briefing

Geopolitics

Washington Wants to Ban China's AI. The Bill Would Land at Home.

Three US agencies accused six Chinese labs of "distilling" American models, and a Senate bill would lock them out of every federal contractor. But cutting off the cheapest AI of the decade would not erase its cost, only move it, and the advisory's own fix asks American firms to quietly serve worse answers to paying customers.

Read full story →
AI Safety

From 'We Might All Die' to 'Let's Coordinate'

A week after their engineers went public with extinction fears, the bosses of Anthropic and OpenAI have a plan: pace the race, install outside auditors, and maybe agree release timing with rivals. That last idea looks uncomfortably like the thing antitrust law exists to stop, and a China-focused White House is not signing up.

Read full story →
AI Infrastructure

OpenAI Sold Its Best Plan So Well It Had to Stop Selling It

OpenAI froze new sign-ups for its $200 Pro tier because its new agent model burns compute faster than money can buy it. In the agent era, one user can set dozens of agents running at once, and the real bottleneck is turning out to be electricity and silicon, not how clever the model is.

Read full story →