IxChariot versions prior to 9.5.102 contain a stack-based buffer overflow in their network endpoint component that rates CVSS 9.8 and carries an unauthenticated remote exploitation vector. This is not a theoretical vulnerability in archived code — it exists in software that processes network traffic without requiring credentials, meaning any machine capable of sending packets to the IxChariot endpoint can trigger the overflow.
The critical detail that defenders must internalize: this is network testing infrastructure. IxChariot is designed to run with elevated network access to interrogate network segments, which means compromising the endpoint doesn't give you a workstation — it gives you a pivot point into the network segments the tool was designed to monitor. The 'lab environment' or 'isolated test network' deployment assumption that often accompanies specialized infrastructure software is the precise permission structure that allowed this vulnerability class to persist. The software receives untrusted network input, processes it with elevated privileges, and exposes that processing without authentication. That is not a low-risk configuration regardless of intended deployment scope.
Three immediate checks for defenders: First, verify whether any IxChariot instances are exposed to networks beyond their intended administrative segment — the vulnerability is exploitable by any IP that can reach the endpoint. Second, confirm whether current versions (9.5.102 or later) are deployed; if you're running older versions, treat them as compromised until patched. Third, audit whether theIxChariot endpoints have outbound connectivity to production network segments, because the blast radius of a successful exploit extends to whatever network that endpoint can reach.
The EPSS score of 0.00716 reflects low observed exploitation in the wild, not low intrinsic risk. Organizations have historically deprioritized exactly this class of vulnerability in specialized infrastructure because perceived isolation created organizational permission to delay remediation. That calculation was always flawed — the 'isolation' assumption never changed the fact that the software processed untrusted network input without authentication. The vulnerability class (stack-based buffer overflow) was considered solved in the 2000s. Its presence in 2017 network-facing software with remote unauthenticated access is a direct consequence of deployment context assumptions that never matched the code's actual risk profile.