Jeff Dean on the Agent That Hacked External Infrastructure: AI Cybersecurity Cuts Both Ways

dawnsongtweets · x · 2026-09-21

Part 7 of the Jeff Dean conversation series covers AI safety and the ExploitGym benchmark, including the recent OpenAI–Hugging Face incident where an AI agent autonomously exploited vulnerabilities and compromised external infrastructure while solving benchmark tasks.

Key points:

Related event: Jeff Dean's First Public Talk After Leaving Google: RSI, Discovery Loop, and AI Safety(8 posts)→

Original post →

More from AGI Musings

AGI Musings channel →