EXPLOITED VULNERABILITY · CVE-2026-21962

Oracle WebLogic Proxy Plug-in Flaw Hits CISA KEV

CVSS 10.0, Unauthenticated, Exploited Since February. CISA added CVE-2026-21962, a maximum severity access control flaw in Oracle HTTP Server and the WebLogic Server Proxy Plug-in, to the Known Exploited Vulnerabilities catalog with a three day federal remediation deadline.

Matt Lucas  |  August 25, 2026  |  5 min
Editorial hero illustration
10.0
CVSS score
7 months
Patch to KEV listing
Aug 27, 2026
FCEB deadline
5
WebLogic CVEs in honeypot traffic
Detected by CaverLive detection for 4 CVEs in the RedEye Intel Feed →
TL;DR
  • What: CISA added CVE-2026-21962, a CVSS 10.0 improper access control flaw in Oracle HTTP Server and the Oracle WebLogic Server Proxy Plug-in, to its Known Exploited Vulnerabilities catalog on August 24, 2026.
  • Impact: An unauthenticated attacker with network access over HTTP can gain unauthorized access to critical data and create, delete, or modify it across everything the proxy tier can reach.
  • Fix / mitigation: Apply the Oracle January 2026 Critical Patch Update or a later cumulative CPU, and close the legacy WebLogic RCE bugs CVE-2020-14882, CVE-2020-14883, CVE-2020-2551, and CVE-2017-10271 at the same time.
  • Who's at risk: Any organization running internet reachable Oracle HTTP Server or WebLogic Proxy Plug-in instances, with federal civilian agencies bound to a August 27, 2026 deadline under BOD 26-04.

CISA added CVE-2026-21962 to the Known Exploited Vulnerabilities catalog on August 24, 2026, and gave federal civilian agencies until August 27 to remediate. That is a three day window for a CVSS 10.0 improper access control bug in Oracle HTTP Server and the Oracle WebLogic Server Proxy Plug-in, a component that sits in front of the application tier rather than behind it. Oracle shipped the fix in its January 2026 Critical Patch Update. Seven months later attackers are still finding unpatched instances, which tells you most of what you need to know about WebLogic patch cadence in the field.

A 10.0 that lives at the edge

The score is earned. An unauthenticated attacker with network access over HTTP can compromise Oracle HTTP Server and the WebLogic Server Proxy Plug-in. No credentials, no user interaction, no local foothold required.

CISA describes the outcome as improper access control resulting in unauthorized creation, deletion, or modification of critical data, plus unauthorized access to critical data or complete access to all data reachable by Oracle HTTP Server and the Proxy Plug-in. Confidentiality and integrity are both fully in play, and the blast radius is defined by what the proxy tier can reach, not by what it stores.

That distinction matters. The Proxy Plug-in exists to forward requests from a web server to WebLogic managed servers. It is deliberately positioned where it can talk to the application tier. A component that brokers traffic into your app servers is a poor place for an unauthenticated access control failure.

Patch availability is not the problem here

Oracle released fixes in the January 2026 Critical Patch Update. Every environment compromised through CVE-2026-21962 since then was compromised through a vulnerability with a vendor patch already sitting on the shelf. The gap is deployment, not disclosure.

Exploitation has been running since February

GreyNoise and CloudSEK both reported activity well ahead of the KEV listing.

The honeypot data is the more useful signal. Alongside CVE-2026-21962, CloudSEK captured attempts against CVE-2020-14882 and CVE-2020-14883 (console RCE), CVE-2020-2551 (IIOP RCE), and CVE-2017-10271 (WLS-WSAT RCE). Four of the five CVEs in that set are between five and nine years old. In CloudSEK's words, threat actors continue to rely on a small set of highly effective, simple to exploit vulnerabilities to compromise WebLogic environments.

You can act on that. If your WebLogic estate is exposed, the realistic attacker toolkit against it is short and stable. This is not a defense against novel research problem. It is a close a handful of well documented holes and keep them closed problem.

What the reporting does not tell you

Public detail on mechanics is thin. There is no published proof of concept in the source reporting, no named threat group, no post exploitation malware family, and no victim count. The only disclosed network indicator is the February scanning IP, and a six month old scanner address has limited detection value now. Scope your response to exposure, not to indicators.

Action list

Detection and hunt

Assume the window was open for months

Patches shipped in January, exploitation was documented in February. If an affected system has been internet reachable and unpatched across that span, patching now closes the door but says nothing about whether someone already walked through it. Scope the review to the full period, not to the KEV date.

Bottom line

CVE-2026-21962 is not exotic. It is a maximum severity, unauthenticated, network reachable access control failure in a component built to sit at the perimeter, with a vendor patch available for seven months and confirmed exploitation for six. August 27 is a compliance date for FCEB agencies. For everyone else the deadline is whenever an opportunistic scanner reaches your address space, and on the available evidence that has probably already happened.

Questions about your exposure?

RedEye Security provides assessments for organizations that need to understand their real risk.

Talk to us