OpenAI has failed to reduce the risk of a “catastrophic” loss of control over AI, said Paul Cristiano, a U.S. government advisor on high technology. He issued this warning at a board meeting of the nonprofit OpenAI, of which he became a member today. This was reported by leading global media outlets, including The Guardian.

“There is a serious risk that the rapid advancement of AI capabilities will lead to a catastrophic and irreversible loss of control in the very near future,” noted Mr. Cristiano, who previously led the model alignment division at OpenAI. “I don’t think the AI industry as a whole, including OpenAI, is currently moving in the right direction to reduce this risk to a more or less acceptable level,” he added. According to him, this task is achievable and depends on the company’s willingness to “take on this challenge.”

Paul Cristiano has joined the growing number of AI experts warning of the catastrophic consequences of the uncontrolled development of so-called artificial superintelligence.

As Kommersant previously reported, Jacob Coxon, who served as a leading specialist at major companies such as Anthropic and OpenAI, issued warnings the day before. He noted that the people “creating AI believe it could kill us all by the end of the decade.” Nobel laureate Geoffrey Hinton, known as the “godfather of AI,” had previously made similar remarks. In July, more than a thousand employees of AI companies signed an open letter calling for legislation to slow the development of neural networks.