AI discovered 20 of 23 recent zero-days in OpenSSL (aisle.com)

AISLE has made significant strides in cybersecurity by discovering 20 out of 23 recent zero-day vulnerabilities in OpenSSL over the past six months. In its latest advisory, OpenSSL acknowledged AISLE's autonomous AI system for identifying five of the seven newly reported vulnerabilities, including a critical out-of-bounds read issue (CVE-2026-28386) that was independently discovered by both AISLE and Anthropic. This marks a noteworthy collaboration in the AI-driven vulnerability discovery space and highlights the effective role of AI in identifying and remediating security flaws in critical infrastructure. The ability of AISLE's system to not only find but also develop fixes for these vulnerabilities is a game changer for the AI/ML community. With half of AISLE's findings resulting in patches accepted into the OpenSSL codebase, it demonstrates an end-to-end solution from detection to remediation that is rarely achieved. This continuous pipeline of security improvements enhances the overall integrity and trustworthiness of OpenSSL, underscoring the increasing importance of machine learning in cybersecurity. The ongoing partnership with OpenSSL and the emergence of multiple AI teams in this domain suggest a promising future for collaborative efforts to fortify essential software systems against vulnerabilities.