What happened
David Robinson, who worked at OpenAI preparing the safety reports published alongside every major model release, resigned this week. He laid out his criticism of the industry's approach to AI safety in an editorial for The Atlantic.
The core argument
According to Robinson, the problem is not something that can be fixed by adding a few new rules or regulations. He argues that the industry's safety culture is fundamentally broken — meaning the issue goes beyond procedures to how companies actually treat safety.
Why it matters
It is easy to be cynical when people who helped build these systems later warn about them. But that does not mean their warnings should be dismissed. An insider's account of how the industry handles safety is worth taking seriously.
