CVE-2026-114835: An Operating Model for a Critical Console Fix

An emergency patch becomes easier to defend when the work follows a repeatable operating model. CVE-2026-114835 is an Oracle Access Manager Admin Console vulnerability in versions 12.2.1.4.0 and 14.1.2.1.0. It has a CVSS 3.1 score of 10.0, an unauthenticated HTTP path, and takeover impact. The November 2026 CSPU carries the fix. The response should combine patch execution, access reduction, validation, and a clear decision record.

Vulnerability overview

Oracle's risk matrix names the product and component directly. The attack is remotely reachable without authentication, with low complexity and no user interaction. The score includes a scope change and high confidentiality, integrity, and availability impact. The recorded result is takeover of Access Manager.

The CVE is distinct from CVE-2026-91500 and CVE-2026-91510 in the Authentication Engine. An operating model that groups all Access Manager findings into one undifferentiated task will lose those boundaries and weaken validation.

Mechanism and exploitation conditions

Oracle does not publish the exact root-cause code. The advisory and CWE-306 enrichment support a missing-authentication condition in a critical administrative function. Exploitation depends on an affected, reachable console. That condition should be tested through the same network paths an unauthorized user could use.

Do not confuse a reachable login page with proof that the vulnerability is exploitable. The public material does not provide a test request. Patched-state verification should use vendor records and supported checks, not a homegrown exploit against production.

Impact

Takeover of the identity administration service can affect policy, trust, and dependent application access. A scope change expands the potential consequence beyond the vulnerable component. The exact blast radius is deployment-specific and should be mapped from Access Manager's integrations and administrator roles.

The reviewed sources do not confirm active exploitation. That gap belongs in the decision record, but it does not justify waiting for an incident before patching. A low-complexity, unauthenticated takeover flaw is already a priority condition.

Affected products and scope

Include both supported affected versions and every deployed node. The Admin Console may be associated with the WebLogic AdminServer and port 7001 by default, so inventory should record the true listener and all front-end paths. Add the application owners, network owners, and administrators who can reach each console.

Define a complete unit of work. One unit should include the Access Manager domain, all nodes in the domain, patch installation, service restart, post-patch verification, and removal or review of temporary access changes. Disaster recovery and lower environments belong in the same plan when they can receive production traffic.

Exposure context

The ZoomEye query title="Oracle Access Manager" returned 1,033 matches at the recorded time. The app and product searches returned zero, and the CVE filter returned zero. Use the title count as a discovery prompt, not as proof that an asset is vulnerable or unpatched.

The stronger control is an internal inventory of console URLs, versions, patch levels, and network paths. External observations can identify unknown owners for follow-up, while internal records determine whether remediation is complete.

Remediation and mitigations

Apply the Oracle November 2026 CSPU through the organization's support and change process. Verify the installed patch and service build, restart the affected processes, and test console access and dependent identity flows. Record who approved any temporary delay, what compensating network control is in place, and when the exception will be reviewed.

Restrict the Admin Console to trusted management paths and review proxy and firewall rules for overlooked routes. Monitor administrative authentication and policy changes. These steps reduce exposure while patching is underway, but Oracle does not list a CVE-specific workaround. The operating model closes only when the update is verified on every affected node.

References