Azure Managed Instance for Apache Cassandra EUVD-2026-28451

Requires authenticated access with low privileges (PR:L) to Azure Managed Instance for Apache Cassandra combined with mandatory user interaction (UI:R) from an authenticated user or administrator. The attacker must successfully deliver malicious input that an authenticated user processes through the management interface, query console, data import functionality, or another service component with inadequate input validation. Network access (AV:N) enables remote exploitation, but the UI:R requirement means the attack cannot be fully automated - successful exploitation depends on social engineering, phishing, or compromising applications that legitimate users interact with while authenticated. The low attack complexity (AC:L) suggests no specialized timing, race conditions, or environmental dependencies beyond the core authentication and interaction requirements.

Real-world risk is elevated but nuanced across multiple signals. CVSS 9.0 Critical score driven by complete confidentiality, integrity, and availability impact combined with scope change indicating potential cloud tenant isolation breach. However, exploitation requires authenticated access (PR:L), user interaction (UI:R), and network positioning (AV:N), significantly raising the attack complexity bar in practice. No EPSS score provided, but absence from CISA KEV and no public POC suggests limited real-world exploitation to date. The UI:R requirement is particularly important - attackers cannot exploit this remotely without tricking an authenticated administrator or user into interacting with malicious input, substantially reducing automatable mass-exploitation risk. CVSS temporal metrics (E:U/RL:O/RC:C) confirm no known exploits exist and official remediation is available with complete vendor confidence. Primary risk scenarios: targeted attacks against high-value Azure Cassandra deployments, supply chain compromises delivering malicious data to legitimate administrators, or insider threats with existing low-privilege access. This is a legitimate critical-severity issue requiring prompt patching, but not an internet-wide wormable threat.

An attacker with low-privilege authenticated access to an Azure Cassandra instance crafts malicious input exploiting inadequate validation in the management interface or data ingestion pipeline. They deliver this payload through a phishing email or compromised application that tricks an authenticated administrator into importing malicious data, executing a crafted query, or clicking a specially-formatted link while logged into the Azure portal. The improper validation allows the attacker's code to execute with elevated privileges, breaking isolation boundaries (scope change) to access other Azure resources, exfiltrate data across tenant boundaries, or establish persistence in the cloud environment.