OpenAI Safety Lead Quits, Says Its Culture Is 'Broken'
The Facts
- David Robinson, who led the writing of safety reports accompanying OpenAI's major product releases, resigned and published an essay in The Atlantic on Saturday titled "I quit OpenAI because its culture is broken."
- Robinson said he spent "three and a half years at OpenAI," making him one of its "longest-tenured employees." He led the drafting of the company's current "Preparedness Framework" and "oversaw the writing of safety reports on 12 frontier launches."
- In the essay, he wrote that OpenAI relies on "trial and error," which it calls "iterative deployment," and said that approach "guarantees periodic failures — and the scale of those failures is growing as systems get more capable."
Sources Split
LEFT
CENTER
RIGHT
+2
+1
+2
+3
ESTAB
CRIT
CENTER
PRO
ESTAB
The Spin
Industry-critical narrative
An industry that treats catastrophic risk as a bug to be patched after launch is gambling with systems nobody fully understands. Humility, outside expertise and layered redundancy of the kind demanded in nuclear plants and air traffic control should be the baseline before the next frontier system ships, not after. Endless sprints and unimpeded optimism are a terrible way to raise minds that may one day outthink their makers.
Pro-industry narrative
Serious safety work is already happening inside the labs, and it looks nothing like reckless sprinting. Labs use layered defenses such as structured safety cases before training runs, leadership veto power, containment red-teaming, live monitoring with auto-pause and public postmortems. Slowing releases when the evidence demands it is a commitment worth judging by results, not by resignation letters.
