CVE-2026-26740 is a vulnerability in Red Hat AI Inference Server 3.2.2 (CUDA) identified by Red Hat Product Security. The issue is associated with multiple CWEs such as CWE-787 (out-of-bounds write), CWE-191, CWE-130, and others, indicating potential memory corruption and related security weaknesses. The CVSS vector indicates network attack vector, low attack complexity, no privileges required, no user interaction, unchanged scope, no confidentiality impact, low integrity impact, and high availability impact. The vulnerability affects the integrity and availability of the AI Inference Server. Red Hat has published advisories referencing this CVE but has not provided a patch or fix as of the advisory date. The affected product includes Red Hat AI Inference Server 3.2 and specific container images for amd64 and arm64 architectures. There are no known exploits in the wild, and the vendor advisory does not indicate any mitigation or workaround.

The vulnerability impacts the integrity and availability of Red Hat AI Inference Server 3.2.2 (CUDA). According to the CVSS vector, the attack can be performed remotely without privileges or user interaction, potentially causing a high impact on availability and a low impact on integrity. There is no confidentiality impact. No known exploits are reported in the wild. The lack of an official fix means the vulnerability remains unmitigated in affected versions.

As of the latest Red Hat advisory, no official fix or patch is available for CVE-2026-26740. Users should monitor Red Hat's security advisories for updates regarding patches or mitigations. Since no workaround or temporary fix is provided by the vendor, cautious deployment and limiting exposure of the affected AI Inference Server instances to untrusted networks may reduce risk until a fix is released.