Former OpenAI Safety Specialist David Robinson Resigns, Raises Concerns Over AI Risk Management

David Robinson, a former safety specialist at OpenAI, has resigned after three and a half years at the company, citing concerns about its organisational culture and approach to managing artificial intelligence risks.

In an essay published in The Atlantic titled I Quit OpenAI Because Its Culture Is Broken, Robinson argued that the company’s focus on rapid product development and frequent launches can leave safety considerations struggling to keep pace with technological progress.

Robinson questioned the industry’s reliance on iterative deployment, an approach that involves releasing AI systems, monitoring their performance and addressing problems as they emerge. He argued that increasingly capable AI systems require stronger safeguards before deployment, drawing comparisons with safety standards in aviation and nuclear power.

During his time at OpenAI, Robinson contributed to the development of its preparedness framework and oversaw safety reports for 12 frontier AI model launches. However, he said the organisation’s repeated focus on launching new models had made it difficult to achieve the level of safety oversight he considered necessary.

OpenAI, in response to concerns about its safety practices, has said it pauses training or delays the release of models when significant safety issues arise. The differing perspectives highlight an ongoing debate within the AI industry over whether existing safety measures and alignment research are keeping pace with advances in AI capabilities.

What Robinson’s Departure Means for HR Leaders

Beyond the technical questions surrounding AI safety, the departure raises important issues about organisational culture, employee voice and leadership accountability.

In high-stakes industries, employees responsible for risk management must be able to raise concerns and influence decisions without feeling that commercial priorities consistently take precedence. When safety specialists believe their expertise is not adequately considered, organisations may face challenges involving employee trust, engagement and retention.

For HR leaders, the situation underscores the importance of establishing clear escalation channels, protecting employees who raise legitimate concerns and ensuring that risk management teams have meaningful involvement in strategic decisions.

As AI companies continue to compete on innovation and deployment speed, maintaining a balance between commercial growth and responsible governance will remain a significant leadership challenge.

Robinson’s departure adds to the broader discussion about how AI companies can sustain innovation while building workplace cultures that support transparency, accountability and effective risk management.