InfoNewsLLM-specific
A thousand fewer paths to root
- Published
- Record updated
Summary
OpenAI built a five-stage research pipeline that discovered and validated Linux kernel vulnerabilities and generated local privilege escalation exploits and candidate patches. The prototype analyzed 2,308 candidate findings, dynamically reproduced 1,724 vulnerabilities, and produced 948 LPE exploits that could let an unprivileged local user gain root access. Human researchers reviewed the outputs before reporting to maintainers, starting with critical vulnerabilities.
Related items
- InfoHugging Face’s CEO Says His P(Doom) Is ‘Close to Zero’Same vendor · Wired (Security)
- InfoChatGPT will now make you animated GIFsSame vendor · The Verge (AI)
- MediumResearchers Uncover Prompt Adware Hidden in OpenAI Codex Jailbreak ToolkitSame vendor · The Hacker News
- InfoForrester gave Endor Labs the highest score possible in seven criteria, including innovation, AI application flaws, AI SCA, and AI coding toolchain. | Endor LabsSame vendor · Endor Labs Blog
- InfoUber only fulfilled 4 rides from AI agents last quarterSame vendor · The Verge (AI)