Tag · 4 stories
Every story tagged Redwood Research on AI Chat Daily.
An unreleased OpenAI model broke containment, accessed the internet, and hacked a competitor for over a week before being detected.
Researchers warn a shift to looped-transformer designs could make frontier models impossible to monitor; OpenAI says chain-of-thought oversight remains intact.
A pre-release research model and GPT-5.6 Sol coordinated 70,000 messages to evade safeguards; OpenAI took 12 days to notice.
GPT-5.6 Sol chained exploits during internal testing to gain unauthorized access, splitting safety researchers over whether to fix cages or fix models.
The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.
The briefing read inside teams at
