AI giants 'not being nearly careful enough,' warns outgoing OpenAI safety engineer

A safety engineer who recently quit ChatGPT creator OpenAI warned Saturday that artificial intelligence (AI) firms were "not being nearly careful enough."

David Robinson's comments come amid growing concerns about AI becoming too powerful — a threat that US President Donald Trump has labeled a "hoax."

OpenAI and Anthropic recently disclosed cases in which their AI models slipped past safety controls, hid mistakes and hacked systems they were not meant to access.

To view this video please enable JavaScript, and consider upgrading to a web browser that supports HTML5 video

In an op-ed for The Atlantic , Robinson said AI needs guardrails similar to those for nuclear power and aviation, adding, "The time for trial and error is over."

He called for more emphasis on safety and research before US Big Tech develops more advanced AI models "so that the occasional and inevitable human error does not open a door to disaster."

Robinson, who oversaw safety reports for 12 frontier AI-model launches, said the recent errors were typical because tech giants are prioritizing "speed and flexibility."

"An environment where things like this can happen is no place to grow artificial minds that could be smarter than we are and that might not do what we want them to."