Safeguard
Vulnerability Analysis•

CVE-2023-25925: High Command Injection in IBM Guardium GKLM

CVE-2023-25925 lets a remote authenticated user run commands on IBM Guardium Key Lifecycle Manager 3.0 to 4.1.1. Upgrade to 4.2 or fix pack 4.1.1 FP7.

Safeguard Security Team
3 min read

CVE-2023-25925 is a high severity (CVSS 8.5) OS command injection in IBM Security Guardium Key Lifecycle Manager 3.0, 3.0.1, 4.0, 4.1 and 4.1.1. A remote authenticated attacker could run commands with a crafted request. IBM lists v4.2 and fix pack 4.1.1 FP7 as the fixes.

What's affected

Both the record and IBM Security Bulletin 6964516 list GKLM 3.0, 3.0.1, 4.0, 4.1 and 4.1.1. IBM says the issue is fixed in v4.2 and v4.1.1.7 and recommends upgrading to v4.2 or applying the latest fix pack, 4.1.1 FP7.

The bulletin gives a base score of 8.5 for this CVE, with a network vector, high complexity, low privileges, no user interaction and a changed scope. The record agrees at 8.5. IBM's X-Force ID is 247632. I found no disagreement on severity or versions.

The weakness is CWE-78, improper neutralization of special elements in an OS command. IBM's description says only that a remote authenticated attacker could execute arbitrary commands on the system by sending a specially crafted request. I will not go further than that, and the advisory does not either.

How to check if you're exposed

IBM Security Guardium Key Lifecycle Manager (GKLM) is an appliance style server application, so the version is read from the product itself. Open the GKLM web console, or follow IBM's documentation for your platform, and note the version and fix pack level. The record also lists AIX among the affected platforms, so the same applies to GKLM installations on that operating system.

Compare what you see against the bulletin:

  • Versions 3.0, 3.0.1, 4.0, 4.1 and 4.1.1 are in the affected list.
  • The fixes are GKLM v4.2, or fix pack 4.1.1 FP7 (version 4.1.1.7) on the 4.1.1 line.

I did not find a documented one line command in the sources I read, so I am not offering one. If you cannot see the fix pack level, ask your IBM administrator to read it from the console or from the install records.

To see which of your own applications pull in affected components, run the free in-browser dependency scan at gold.safeguard.sh/scan.

What to do now

  1. Find all GKLM servers and record the version and fix pack.
  2. Upgrade to v4.2, or apply fix pack 4.1.1 FP7 on the 4.1.1 line.
  3. Back up configuration and key material before changing anything.
  4. Audit which accounts have console or API access and remove the ones you do not need.
  5. Review the server and its logs for commands or processes you cannot explain.

Why it matters

Command injection turns data into instructions: input meant to be a value ends up interpreted by the operating system. In a product that stores encryption keys, running commands on the host is close to the worst case, because the host holds secrets for other systems.

The attacker must be authenticated, which is why this does not score 9 or above. Weak or shared admin credentials erase that protection, though. The record does not list the CVE in CISA KEV, and the bulletin makes no statement about exploitation.

Quick check

Is login required?

Yes. IBM describes a remote authenticated attacker and the record shows low privileges required.

Where do I read the fix?

IBM Security Bulletin 6964516 lists v4.2 and fix pack 4.1.1 FP7.

Sources

Last reviewed: 19 October 2026.

Never miss an update

Weekly insights on software supply chain security, delivered to your inbox.

We email you a link to confirm, and only then add you to the list. Unsubscribe in one click. Privacy

Self-healing security runs on Safeguard.

Your first fix PR is minutes away.

No sales call required, even your agent can complete the purchase over MCP.