CVE-2026-104742

AI Puffer <= 2.4.89 - Missing Authorization to Authenticated (Subscriber+) Semantic Search Settings Modification via ajax_save_semantic_search_settings AJAX action

2026-10-09 17:12
Wordfence PRISM

Strategic Overview

Status
Patched in 2.4.90
Affected Version
<= 2.4.89
CVSS
3.1Low
Weakness type
CWE-862 · Missing Authorization
CVE
CVE-2026-104742
View all AI Puffer – MCP, AI Chatbot & Content Automation vulnerabilities

At a glance

CVE-2026-104742 is a low-severity Missing Authorization vulnerability in the AI Puffer WordPress plugin, affecting versions <= 2.4.89. It carries a CVSS score of 3.1 (reachable over the network). Exploitation requires an authenticated account at Subscriber level or above. The issue is fixed in version 2.4.90; sites on affected versions should update now. Disclosed October 2026, reported by Wordfence PRISM.

Vulnerability Overview

The AI Puffer – Chat. Create. Automate. (formerly AI Power) plugin for WordPress is vulnerable to authorization bypass in all versions up to, and including, 2.4.89. This is due to the plugin not properly verifying that a user is authorized to perform an action. This makes it possible for authenticated attackers, with subscriber-level access and above, to modify global site-wide semantic search settings, including vector provider, embedding provider, embedding model, target ID, number of results, and no-results text stored in aipkit_options, that are otherwise restricted to administrators. Exploitation requires that an administrator has previously granted the Knowledge Base ('sources') module to the attacker's role via the Role Manager, as this access is not available to lower-privileged users by default.

Technical Analysis

The vector marks this flaw as remotely reachable over the network, and no interaction from a victim user.

CWE-862: Missing Authorization

AI Puffer <= 2.4.89 carries this weakness at sources, and reaching it takes an account at Subscriber level or above. A missing authorization check means a function is reachable by anyone who can reach the endpoint, because the code never asks whether the caller is allowed to perform the action.

Any user who can reach the endpoint gets to run an action reserved for higher-privileged roles, from reading protected data to changing settings or content. For AI Puffer the fix is 2.4.90: builds <= 2.4.89 are affected, anything from 2.4.90 onward is not.

Remediation

Update to version 2.4.90, or a newer patched version

How does WordSec protect against this?

Because it turns on account access, WordSec's login security is the relevant layer: role-based two-factor, captcha and brute-force limits raise the cost of getting the account this needs. None of that substitutes for the fix: AI Puffer 2.4.90 closes this, and updating the plugin is the step that ends it.

  • Login Security
  • Alerts

External References

Related records

Vulnerability data © Defiant, Inc., provided under the Wordfence Intelligence T&C