Oracle Platform Security EUVD-2026-79339

Exploitation requires HTTP network reachability to a service that loads the OPSS security provider from Oracle Fusion Middleware running OPSS 12.2.1.4.0 or 14.1.2.0.0, and no authentication or user interaction is required per the CVSS vector (PR:N, UI:N). The limiting factor is the High attack complexity (AC:H) that Oracle itself characterizes as 'difficult to exploit' - success depends on conditions the attacker cannot fully control, such as a race window, a specific server state, or an exact request form, and Oracle released no technical detail about what those preconditions are. The defect is confined to the bundled 'Centralized Thirdparty Jars' component, so environments that do not load OPSS (or that have removed the OPSS security provider) are not exposed. Note that beyond HTTP reachability the specific triggering configuration is not documented in the available data, and the affected version range is limited to the two OPSS releases named by Oracle.

The signals are mixed and should be read carefully. CVSS 3.1 gives 8.1 (High) with AV:N/PR:N/UI:N, which is the profile of an unauthenticated network attack requiring no user interaction, but AC:H explicitly encodes that the attacker must win a race, hit a specific state, or otherwise satisfy brittle preconditions - Oracle's own wording calls it 'difficult to exploit'. No EPSS score, no CISA KEV listing, and no public proof-of-concept were provided in the input, so there is no evidence of active exploitation and the prioritizing signal is the CVSS vector rather than observed threat activity. Note the tension between the tags and the description: the input tags say 'Information Disclosure', yet the vector assigns High Confidentiality, Integrity AND Availability impact and the description states 'takeover of Oracle Platform Security for Java' - these are contradictory and the description/vector should be treated as authoritative over the tag. The missing CWE and missing EPSS data are explicit gaps; treat the impact as full compromise (including potential authentication and policy bypass across anything that trusts OPSS) but do not treat this as an emergency without confirmation from Oracle's advisory.

An unauthenticated attacker with HTTP reachability to an OPSS-exposed Fusion Middleware endpoint sends a specifically crafted request that triggers the flaw in the centralized third-party jar, but the attacker must satisfy a non-trivial precondition (AC:H) - likely a timing window, a particular server state, or a specific request shape - to succeed. On success the attacker gains control over OPSS itself, which is trusted by the surrounding middleware, allowing policy, credential-handling and application code paths to be subverted for full compromise. No public exploit code has been identified, so this scenario currently requires bespoke research rather than a copy-paste tool.