OpenAI’s AGI Safety Lead Resigns, Warns of Unprepared Future

OpenAI’s AGI readiness czar, Miles Brundage, has resigned, signaling a significant shift in the company’s leadership as it pursues artificial general intelligence (AGI). Brundage, who led efforts to ensure AGI’s safe integration, warns that neither OpenAI nor the world is adequately prepared for the transformative power of AGI.

In a recent post on his Substack, Brundage shared that leaving his “dream job” after six years was challenging, but he believes his influence on AI’s future will be stronger from outside the organization. He stated, “I want to impact and influence AI’s development from outside the industry.” His main concern: AGI might arrive before society is ready to handle its implications.

Brundage’s concerns echo his involvement in OpenAI’s five-tier AGI scale, which he helped develop to gauge AI maturity. The scale, which was leaked to Bloomberg, defines AI’s evolution toward human-level intelligence, with OpenAI currently positioned at level two of five. Brundage emphasized, however, that AGI is more nuanced than a simple “before and after” scenario, suggesting that preparedness at every stage is crucial.

His departure comes months after OpenAI disbanded its AI safety team, a move that raises further concerns about the company’s approach to AGI. Brundage still believes in AGI’s potential for humanity but underscores the need for careful governance. As OpenAI reallocates its AGI readiness team, Brundage’s resignation underscores the urgency of responsible AGI development in an increasingly accelerated field.