OpenAI discovered 948 distinct privilege-escalation vulnerabilities in the Linux kernel and developed working exploits using autonomous security agents, the company announced Wednesday. The research cost $1.21 million in API credits, averaging $1,279 per vulnerability discovered. OpenAI conducted the research to study how effectively frontier AI models could find vulnerabilities in a codebase of Linux's scale and importance, which powers cloud servers, Android devices, and critical computing infrastructure worldwide.
OpenAI Agents Found 948 Flaws and Created 1,169 Exploits
OpenAI's agents scanned approximately 2.8 million lines of kernel code, representing roughly 7% of the Linux project. The prototype identified and analyzed 2,308 candidate findings, dynamically reproduced 1,724 vulnerabilities, and produced 948 local privilege escalation exploits that could allow an unprivileged local user to gain root access. After grouping exploits by their underlying cause and proposed fix, researchers counted 948 distinct flaws. The flaws identified were local privilege-escalation vulnerabilities, meaning an attacker would first need limited access to a vulnerable machine but could then potentially obtain root privileges and take full control.
Research Cost $1.21 Million with Newer Model Finding 5x More Vulnerabilities
The October run cost $1,213,430 in API credits, equivalent to around $1,279 per vulnerability. OpenAI compared that with historical $20,000-to-$70,000 payouts for Linux privilege-escalation exploits at hacking competitions, suggesting a hypothetical value of $18.9 million to $66.3 million if the same results were found manually. An earlier GPT-5.6 Sol campaign found 166 unique privilege-escalation vulnerabilities in July for $513,690. A newer, unreleased cyber model found more than five times as many in October, even after the kernel had undergone months of fixes. The system also generated candidate patches, with human review and software maintainers remaining responsible for validating and implementing fixes.
AI Agents Discovered at Least 10 Novel Exploitation Techniques
OpenAI stated its agents used at least 10 exploitation techniques unfamiliar to the researchers and outside experts they consulted—not merely existing tricks retrieved from training data. The company's autonomous security agents automated much of the exploitation process, creating a potential advantage for defenders trying to patch software before criminals can take advantage of the same weaknesses.
Anthropic Disclosed Similar Vulnerability Findings Last Week
Last week, Anthropic disclosed that an unreleased version of Claude had discovered and exploited serious vulnerabilities in widely used software, including Chrome and Linux. Some of those exploits could execute malicious code remotely with little or no user interaction. Researchers at Vidoc Security reproduced several vulnerability discoveries associated with Anthropic's powerful Mythos AI using publicly accessible models for less than $30 per file. Those tests showed that finding vulnerabilities was becoming increasingly accessible, although reproducing a flaw is different from developing a reliable exploit capable of compromising a protected system.
OpenAI Restricts Access Through Daybreak Red Security Program
OpenAI is attempting to restrict access to its more capable cybersecurity models through verified security programs. Its Daybreak Red initiative provides approved defenders with access to specialized AI tools for vulnerability research, exploit testing, and mitigation development. The company warned that attackers with limited access to an outdated operating system may increasingly be able to escalate their privileges with AI assistance.
FAQ
What did OpenAI discover in the Linux kernel?
OpenAI discovered 948 distinct privilege-escalation vulnerabilities in the Linux kernel using autonomous security agents. The agents scanned approximately 2.8 million lines of kernel code, identified 2,308 candidate findings, dynamically reproduced 1,724 vulnerabilities, and produced 948 local privilege escalation exploits that could allow an unprivileged local user to gain root access.
How much did OpenAI's Linux vulnerability research cost?
The October research run cost $1,213,430 in API credits, averaging approximately $1,279 per vulnerability discovered. An earlier July campaign using GPT-5.6 Sol found 166 unique vulnerabilities for $513,690, while the newer unreleased cyber model found more than five times as many flaws in October.
