- Treasury Secretary Scott Bessent floated a joint U.S.-China notification channel for early warning of AI incidents during pre-summit talks with Vice Premier He Lifeng, ahead of the Trump-Xi meeting. (Joe)
- Details on the proposed channel are thin, but it would cover national security level incidents such as uncontrollable agents and non-state actors in cyber or bio, and the two officials may meet again in Shenzhen in November. (Joe)
- Iran, China and two private Israeli firms used downloadable Chinese open models to run networks of fake accounts pushing false and inflammatory political content on major social platforms. (Joe)
- The campaigns were crude and caught quickly by U.S. intelligence, but they show manipulation can now be delegated to AI and repeated at a scale human staff could not match. (Joe)
- Mitch explains how Effective Altruism, a loose movement combining moral aims with cost-effectiveness reasoning, became a target for the Trump administration, even though most of its members are neither anti-technology nor trying to halt AI. (Mitch)
<- Back to digests
Daily digest
Hold music
By Joe Rogero and Mitchell Howe
U.S.-China hotline talks, the Effective Altruist boogeyman, and more
Read the full dispatch on AI StopWatch
From AI StopWatch, published with their permission.
The summaries and the full Spanish translation are produced automatically. The original is always linked.