AI pipeline finds hundreds of linux kernel flaws, demonstrating automated patching potential

A new autonomous pipeline was developed to discover, exploit, and patch vulnerabilities within the Linux kernel. In a single run using an upcoming cyber model in mid-October 2026, the system identified and analyzed 2,308 candidate findings, dynamically reproduced 1,724 vulnerabilities, and generated 948 local privilege escalation (LPE) exploits for a common Linux distribution. This process incurred an API cost of $1,213,430. This research demonstrated the pipeline's capability across five stages: discovery, dynamic validation, exploitability assessment, patch generation, and human review. The system's output includes a root cause analysis, a minimal crashing proof-of-concept, a candidate patch, and a full exploit. When tested on a reduced attack surface environment (kernelCTF), the pipeline yielded 21 LPE exploits, which, if accepted, could lead to a bounty award of approximately $1.5M. Historically, the 948 LPE exploits found in the main run suggest a hypothetical valuation of $18.9M to $66.3M based on Pwn2Own and TyphoonPWN bounties. The pipeline's methodology, which involved Scout agents identifying 7% of the kernel's code base, suggests that automated patch generation is crucial for securing software at the current pace of vulnerability discovery.