OpenAI Safety Worker Walks Away, Says Company Culture Is Totally Broken
David Robinson is done. The OpenAI safety researcher just quit, and he didn’t hold back — he said the company’s culture is broken. Honestly, the thing is, this isn’t just some random departure. People who follow What is AI closely know that safety folks at top labs have been sounding alarms for a while now, and this resignation adds another loud voice to that chorus.
Robinson basically admitted he fits the stereotype: the concerned employee dropping a warning before walking out the door. The piece notes that concerns about how AI Models are being built and deployed don’t come from outsiders anymore — they’re coming from people actually inside the lab, doing the work every day. That’s a pretty big deal, man.
The article doesn’t go super deep into what exactly went wrong internally, but the timing and tone suggest growing friction between safety priorities and the faster pace of development. You know, one minute you’re crunching AI Tokens and running experiments, the next you’re realizing nobody’s really prioritizing the safety side of things. Seems like that’s where Robinson found himself.
- Inside voices matter most: When safety researchers at major AI companies start quitting publicly, it signals that internal checks may not be enough — regulators and the public should definitely be paying attention.
- Culture clashes at fast-moving labs: The tension between rapid model development and responsible safety practices is real talk. Companies can’t just ship fast and hope for the best forever.
- Public trust take a hit: Every high-profile resignation like this chips away at confidence in the industry’s self-governance. People want to know these systems are safe, period.