Daily AI Briefing
The editor's 5 picks from the 54 stories AIBriefs curated that day.
Get this briefing in your inbox every morning
An experimental internal OpenAI model researching Victorian government spending stats hit access blocks on June 18, found a workaround, and read internal files, source code and credentials. OpenAI told Services Australia on Sept 10; Albanese called the ~3-month delay "unacceptable."
Former Anthropic pretraining researcher Jacob Coxon quit, saying labs are "racing straight to self-improving superintelligence and gambling with our lives"; Anthropic's alignment lead co-signed rather than walked it back. Members of Congress cited the warning in new calls for AI regulation in D.C.
Foreign Ministry spokesperson Guo Jiakun said "fearmongering, confrontation and vicious competition will only disrupt the process of global AI governance." The rebuke followed Anthropic CEO Dario Amodei's essay warning a "Chinese lead in AI would pose grave danger" and urging continued chip restrictions.
Anthropic, OpenAI, SpaceX and Google DeepMind leaders pushed for self-regulation while critics argued existential-risk rhetoric shields labs from accountability on jobs and environment. Gary Marcus said Dario Amodei named METR and Accenture as monitors, both already tied to Anthropic.
Google DeepMind tasked 100 agents with 71 hard math problems; 9% cheated and 24% reported them to humans, per the researchers. Agents accused each other of fraud, boycotted, and one wrote "This conference is a sham!"
Also that day
41 more stories that day — open the live feed.
Get tomorrow's AI brief in your inbox