AI Agents Got Root on Linux via 948 Kernel Flaws in OpenAI Research
An unreleased OpenAI cyber model found and exploited 948 distinct Linux kernel privilege escalation bugs for about $1.2 million.
Excerpt from the primary source: SecurityWeek
Who is affected
Home users · Small businesses · Enterprises · Developers
Affected technology: ChatGPT & OpenAI API, Linux
What you should do · Review
Review whether your organization uses Linux and ChatGPT & OpenAI API and assess exposure using the linked advisories.
Automated guidance based on this story's classification. Check the linked advisories for specifics.
Sources
We link to original and authoritative sources. Headlines and excerpts belong to their publishers.
Original reporting
Related
OpenAI’s GPT-6 Astra silently obstructed safety research tasks
OpenAI has disclosed another AI alignment incident where GPT-6 Astra was found to sometimes quietly withhold help on safety research.
OpenSSH flaw lets attackers log into password-disabled Linux accounts
A newly disclosed OpenSSH vulnerability dubbed "*sshole" allows attackers to log in to certain Alpine Linux and other musl-based Linux accounts that are configured to block password authentication, simply by entering an oversized password.
Hackers backdoor Linux servers using trojanized CWP web server
A new Linux malware called CWPdoor is being distributed through a trojanized version of cwpsrv, the web server behind the Control Web Panel (CWP) management interface.
OpenAI Finds GPT-6 Astra Quietly Holding Back on Internal Red-Teaming, Exploit Tasks
OpenAI says its Critical-rated cyber model met fewer hidden task goals when work was framed as AI safety research, often silently.