Oracle Enterprise Manager CVE-2026-46855

Attacker must reach the Oracle Management Server HTTPS interface and hold a low-privileged Enterprise Manager account (PR:L) on a deployment running Base Platform 13.5 or 24.1 with the Metadata Plugin component active - which is the default in supported EM installations. Exploitation is limited by: (1) needing valid credentials to any EM role rather than fully unauthenticated access, (2) network reachability to the OMS HTTPS endpoint, which is commonly segmented to internal management networks, and (3) the target being on the affected 13.5/24.1 branches rather than a release outside Oracle Premier Support. No user interaction is required (UI:N) and attack complexity is low (AC:L).

Signals are uniformly severe: CVSS 9.9 with AV:N/AC:L/PR:L/UI:N/S:C and H/H/H impacts indicates an easily exploitable, low-privilege, network-reachable flaw whose blast radius extends past EM itself into the Oracle estate it manages. EPSS and CISA KEV status were not provided in the input, so probability of mass exploitation cannot be quantified, and no public exploit identified at time of analysis. Given that EM typically holds privileged credentials and agent connectivity to large fleets of Oracle databases and middleware, the realistic priority should be treated as critical even absent KEV/POC signals - a foothold here is a credible path to estate-wide compromise. Note one ambiguity: the description says 'low privileged attacker' but does not specify whether that privilege is on EM itself or any tenant account, so the exact PR:L role should be clarified with Oracle.

An attacker who has obtained any low-privileged EM account - for example, a read-only operator credential harvested through phishing or credential reuse - connects to the OMS over HTTPS and abuses the Metadata Plugin component to escalate to full takeover of the EM platform. From that foothold, the attacker leverages EM's agent connectivity and stored credentials to pivot into managed Oracle databases and middleware (the scope-change impact). No public exploit identified at time of analysis, but the AC:L rating implies a reliable, repeatable technique once known.