On this page (6 sections)
CVE-2026-114822 is a critical vulnerability in the RDBMS component of Oracle Database Server. According to the NVD record, it is rated CVSS 3.1 base score 9.9 and affects supported versions 23.4.0 through 23.26.4. The record was published on November 17, 2026.
If you run Oracle Database in that version range, or host applications that depend on it, you should review the vendor's alert and plan patching now. The attacker needs only a low privileged account with the Create Session privilege and network access via Oracle Net.
What was disclosed
The description in the NVD entry says the issue is in the RDBMS component. It is described as easily exploitable by a low privileged attacker who holds the Create Session privilege and has network access via Oracle Net. Successful attacks can result in takeover of the RDBMS.
The description also notes a scope change: although the flaw is in the RDBMS, attacks may significantly impact additional products. The GitHub Advisory Database entry (GHSA-mvp9-h8rp-29fv) repeats the same text and is not independent confirmation. Both point to Oracle's security alert page as the reference.
Reading the CVSS score
The vector is CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:C/C:H/I:H/A:H. In plain terms:
- Network attack vector: the attacker does not need local access to the server.
- Low attack complexity and no user interaction: nothing special has to line up, and no one has to click anything.
- Low privileges required: a normal database account with Create Session is enough, not a DBA account.
- Scope changed: impact can reach beyond the database component itself.
- High confidentiality, integrity and availability impact: data can be read, changed or made unavailable.
The practical meaning is that any compromised or untrusted low privileged database login, for example a leaked application credential, becomes a much more serious problem on an unpatched server.
Who is realistically exposed
Exposure depends on two things: running a version between 23.4.0 and 23.26.4, and whether an attacker can obtain a database account and reach Oracle Net.
- Databases reachable from broad internal networks, partner networks or the internet.
- Systems with many accounts that hold Create Session, such as reporting users, application schemas, or shared service accounts.
- Web applications and agencies' client sites that store credentials for an Oracle backend in configuration files, since those credentials could be a starting point.
- Hosting and managed service providers running Oracle Database for multiple tenants.
Databases only reachable from a small, trusted application tier with tightly controlled accounts have lower practical risk, but the vendor patch is still the correct fix.
How to check your exposure
Start by finding every Oracle Database instance and its version. The following read-only SQL query, run in SQL*Plus or another client as a user allowed to query the view, shows the database version banner. Compare the result with the 23.4.0-23.26.4 range.
SELECT banner_full FROM v$version;Next, list which accounts hold the Create Session privilege, since that is the privilege the advisory names. This read-only query uses a standard data dictionary view and needs access to DBA views.
SELECT grantee FROM dba_sys_privs WHERE privilege = 'CREATE SESSION' ORDER BY grantee;Review the list for accounts that are unused, shared or no longer needed. Finally, confirm from a network point of view which hosts can reach the listener port, so you know who could attempt a connection.
What to do now
- Inventory all Oracle Database Server instances and record their exact versions.
- Read Oracle's security alert to identify the patch for your release and any prerequisites.
- Schedule and apply the vendor patch to every instance in the 23.4.0-23.26.4 range, starting with those reachable from untrusted networks.
- Until patched, restrict Oracle Net access with firewall rules or listener controls so only required application hosts can connect.
- Remove or lock unused accounts and revoke Create Session from accounts that do not need it.
- Rotate database credentials that are stored in application configuration files, especially if there is any sign of leakage.
- Back up databases and test the restore before patching, and test the patch in a non-production environment first.
Detection and monitoring afterwards
The sources do not describe any indicators of compromise, so there is no vendor-published signature to search for. Focus on general database hygiene:
- Enable and review database auditing for logins, privilege grants and changes to accounts.
- Watch for logins from unexpected hosts or at unusual times, and for new accounts or privilege grants you did not make.
- Monitor for unexpected changes to database objects or sudden service interruptions.
- Keep an eye on the Oracle alert page and NVD record for updates, since details such as fixed versions may be added.
If your applications sit in front of the database, regularly checking those sites for unexpected changes also helps. A free website malware scanner gives an outside view, and monitoring can alert you to uptime, certificate and DNS changes. Neither can see inside a database server, so they complement patching rather than replace it. For broader guidance, see our Small business website security guide.
Sources
This article is based on the public sources below. Details can change as vendors and researchers publish more; check the original advisories before acting on version numbers.
