Ex-OpenAI Core Researcher Warns of Severe Underinvestment in AGI Alignment

ShakeelHashim · x · 2026-08-04

Yo, a former core capabilities researcher at OpenAI, has criticized the company's approach to AI safety. He noted that despite long-standing discussions about pivoting research focus towards superintelligence (RSI) alignment and control projects, actual execution has been severely lacking.

Currently, only about 20 out of over 1,000 researchers at OpenAI are dedicated to these critical safety initiatives. He argues that this massive gap cannot be bridged fast enough through hiring alone and requires leadership to forcibly shift priorities away from less mission-critical projects. He urges current researchers to push for this change with their voices and actions, emphasizing the urgency of acting within the remaining limited time window.

Related event: Ex-OpenAI Researchers Warn of Severe Safety Underfunding(2 posts)→

Original post →

More from Companies & People

Companies & People channel →