CVE-2026-4775 is a critical security vulnerability in Red Hat AI Inference Server 3.3.5 (ROCm) identified by Red Hat Product Security. The vulnerability involves multiple weakness types including integer overflow and others related to memory and input validation errors. The CVSS vector suggests an attacker with local access and no privileges can exploit this vulnerability with user interaction, potentially causing high impact on confidentiality, integrity, and availability of the system. The official Red Hat advisory (RHSA-2026:30088) mentions the release of version 3.3.5 but does not provide any fixes or patches for this vulnerability. No exploits are currently known in the wild. The affected product is not a cloud service, so remediation depends on vendor patch releases. The advisory references multiple CVEs but does not clarify remediation status for CVE-2026-4775 specifically.

The vulnerability can lead to complete compromise of confidentiality, integrity, and availability of the affected Red Hat AI Inference Server 3.3.5 (ROCm) system. Given the CVSS vector, an attacker with local access and no privileges can exploit this vulnerability with user interaction, potentially leading to severe system compromise. No known exploits in the wild have been reported, but the critical severity indicates a high risk if exploited.

Patch status is not yet confirmed — check the Red Hat advisory (https://access.redhat.com/errata/RHSA-2026:30088) for current remediation guidance. The advisory does not currently list any fixes for CVE-2026-4775. Users should monitor Red Hat security updates for patches addressing this vulnerability. No vendor-provided temporary mitigations or workarounds are indicated in the advisory.