Escalation of Privilege Vulnerability in Intel AI Reference Models

CVE-2026-21400

What is CVE-2026-21400?

The vulnerability presents a significant risk in Intel AI Reference Models prior to version v3.4.1, allowing unprivileged software adversaries to escalate privileges under certain conditions. This issue could potentially be exploited through local access or by leveraging a privileged user. The attack is characterized by low complexity, requiring minimal interaction from the user. If successfully executed, this vulnerability may severely impact system integrity, confidentiality, and availability, posing a notable threat to overall system security.

References