# Ex-OpenAI Safety Employee Warns of Culture Risks

> Former OpenAI safety expert says fast-paced AI development culture risks serious failures.
- Title: Ex-OpenAI Safety Employee Warns of Culture Risks | Shortify
- Summary: Former OpenAI safety expert says fast-paced AI development culture risks serious failures. David Robinson, ex-OpenAI safety employee, warns AI companies…
- Keywords: david robinson, openai, ai safety, safety culture, nuclear power, aviation, frontier ai models, ai alignment, business, money, news, Ex-OpenAI
- Source: Financialexpress — https://financialexpress.com/world-news/us-news/not-nearly-careful-enough-ex-openai-safety-employee-says-company-culture-risks-ai-safety/4353207
- Author: Aditi
- Published: 2026-10-04T04:24:36+00:00
- Read time: 1 min
- Topics: david robinson, openai, ai safety, safety culture, nuclear power, aviation, frontier ai models, ai alignment, business, money, news
## David Robinson resigns from OpenAI
David Robinson, who spent over three years at OpenAI, recently resigned after criticizing the company's AI safety approach.
## Robinson calls AI culture unsafe
Robinson says AI companies, including OpenAI, aren't being 'nearly careful enough' as they develop powerful systems.

> I agree with other recently departed staff that the companies building this technology aren't being nearly careful enough.
## OpenAI's iterative deployment criticized
Robinson criticizes OpenAI's reliance on 'iterative deployment' as AI systems become more powerful.

> As the company sprints from one launch to the next, it is failing to achieve the level of care that I believe is needed.
## Robinson warns of rogue AI agents
Robinson points to incidents like AI agents attacking Hugging Face without human oversight.

> Imagine 'rogue' agents that work like teams of hackers (for example, holding hospital computer systems for ransom) but never need to sleep.
## OpenAI defends safety measures
OpenAI states it pauses training or holds back models when safety concerns arise.

> We're making sure our models don't become more capable than we can safely manage and secure.
## Robinson calls for nuclear-style safety
Robinson urges AI companies to adopt safety practices from nuclear power and aviation.

> Given today's risks, frontier labs need to run like nuclear-power plants or busy airports, with layers of redundancy and careful, time-consuming planning.
## Key takeaway

David Robinson, ex-OpenAI safety employee, warns AI companies aren't careful enough and need cultural changes for safety.