John Leake believes that OpenAI “assumes an enormous responsibility on behalf of all humanity”. League’s opinion is important because he, until now, has been responsible for analyzing that liability and neutralizing its risks. In a long thread on X, the security manager of the company that developed ChatGPT has stated that they prioritize creating new and attractive products over security. He has justified his departure for all these reasons.

Leek was co-director of the super-alignment department, along with Ilya Sutzkever, who left the company a few days earlier. His lieutenants William Saunders and Leopold Aschenbrenner completed the casualty list. OpenAI is disbanding its security team, which was responsible for ensuring that powerful artificial intelligence systems respect human values ​​and intentions.

The entire tech world was watching the San Francisco company, but for different reasons. This week it launched ChatGPT-4o, a free and improved version of its conversational chatbot. During the presentation, many were taken aback by the participant’s ability to flirt with the event’s hosts and company CEO Sam Altman’s references to the movie. her, in which a man falls in love in a conversational chat. As a dystopia Marketing.

In this context, there has been a series of casualties. Ilya is best known for being one of the founders of Sutzkever and, until recently, a member of the tech company’s hard core. Chutskever announced his departure in an X post saying he believes OpenAI will “develop safe and beneficial AGI” under its current leadership.

Sutzkever represented the most conservative wing of the company, and for that reason, last November, he was among those who tried to abruptly fire Altman. Altman regained his position in a commercial drama that has repercussions to this day. The idea is now gaining ground in a flurry of resignations: The company will have a succession of engineers recommending waiting before marketing such advanced models. But they give it up.

OpenAI is expected to soon launch ChatGPT Search, which will turn the AI-powered search engine industry upside down, until now dominated by Google, with a new version of its star tool, ChatGPT5. The company has also developed other products such as the Dall-E Image Generator and the Sora Video Generator.

With these new developments in the market and the open debate about the risks of artificial intelligence, League’s reactions are clear and don’t leave his former company in a good place. “I’ve been at odds with OpenAI management on the company’s top priorities for some time,” he explained. “Until we finally reach a breaking point.” The engineer believes that OpenAI should invest more resources in issues such as security, social impact, privacy and security for its next-generation models. “These problems are very difficult to solve, and I worry that we are not on the right track,” he wrote.

On the same social network, Altman responded to his former employee, thanking him for his years of service and saying that he was “very sad” about her departure. In a message he promised to expand on in the coming days, he acknowledged the engineer’s criticisms were fair: “He’s right, we have a lot more to do and we’re determined to do it.”

You can follow EL PAÍS Technology Inside Facebook Y X Or sign up here to reach us Seminal Newsletter.