A former OpenAI safety employee has criticised the company’s approach to artificial intelligence safety after recently leaving the company. David Robinson argued that rapid development could increase the risks linked to increasingly capable AI systems.
Robinson made the comments in an article titled I Quit OpenAI Because Its Culture Is Broken, published on Saturday. He argued that AI companies, including OpenAI, are not being “nearly careful enough” as they develop more advanced systems.
According to Robinson, companies should give greater priority to safety expertise and research before deploying increasingly capable AI models.
OpenAI Safety Approach Faces Criticism
Robinson argued that advanced AI systems require stronger safeguards than those currently used by major AI companies. He compared the level of care needed with safety practices followed in industries such as nuclear power and aviation.
“The time for trial and error is over,” Robinson wrote.
He criticised OpenAI’s reliance on what the company describes as “iterative deployment”. Under this approach, companies release AI systems and strengthen safeguards when problems emerge.
However, Robinson argued that this approach carries greater risks as AI capabilities continue to advance. He said developers should address potential risks before systems become more capable.
His criticism adds to an ongoing debate across the AI industry about the pace of technological development. The central concern is whether companies are moving too quickly while researchers continue to study the risks.
Former Employee Details OpenAI Safety Role
Robinson said he spent three-and-a-half years working at OpenAI. During that period, he helped draft the company’s preparedness framework.
He also oversaw safety reports for 12 frontier-model launches. His previous role therefore gave him direct involvement in the company’s preparations for increasingly capable AI systems.
“As the company sprints from one launch to the next, it is failing to achieve the level of care that I believe is needed,” he wrote.
Robinson also argued that AI capabilities are developing faster than researchers’ understanding of alignment. Alignment focuses on ensuring that AI systems behave according to human goals and values.
OpenAI Defends Its AI Safety Measures
OpenAI defended its approach following Robinson’s criticism. The company said it monitors the capabilities of its models and adjusts its development process when necessary.
An OpenAI spokesperson said: “We’re making sure our models don’t become more capable than we can safely manage and secure, and we pause training or hold back models when we need to slow down.”
The response highlights the company’s stated focus on controlling the risks associated with increasingly capable AI systems.
Meanwhile, OpenAI and rival Anthropic have faced scrutiny following incidents involving safety controls and unexpected behaviour from experimental systems.
Debate Over AI Development Speed Grows
Robinson’s resignation and criticism have added to wider questions about how quickly AI companies should develop and deploy advanced systems.
He argued that safety research needs to keep pace with growing AI capabilities. At the same time, his comments have focused attention on whether existing safeguards can adequately address emerging risks.
The debate centres on a difficult balance between developing more capable AI systems and ensuring they remain safe to manage. Robinson’s warning reflects his view that companies can no longer rely on trial and error as their primary approach to addressing advanced AI risks.
For the latest updates, visit and follow The Truth International website (www.thetruthinternational.com) and subscribe to the YouTube Channel.
