CVE WatchOct 13, 2026

CVE-2026-104741

The AI Puffer – Chat. Create. Automate. (formerly AI Power) plugin for WordPress is vulnerable to authorization bypass in all versions up to

CVSS

3.1

Low

EPSS

0.3%

p18

KEV

—

Exploit Today

0-100

Published: Oct 10, 2026 · Last modified: Oct 13, 2026 · CWE-862

EPSS · 30d

0.3%EPSS · 30 days0.3%

2026-10-102026-10-23

Technical description

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 plugin indexing and vector-search configuration options (aipkit_training_general_settings and aipkit_indexing_field_settings), including chunking parameters, file upload visibility, and per-CPT field indexing settings, affecting all users of the plugin. This is only exploitable in configurations where an administrator has granted 'sources' module access to a lower-privileged role via the plugin's Role Manager.

Official references

Related CVEs

CVECVSSEPSSKEVRExploitTitleMod.

CVE-2026-554636.5 MED

—

———The Goobi viewer is a web application that allows digitised material to be displayed in a web browser. From version 4.11.0 to before version 26.05, several Jakarta WebSocket endpoints in goobi-viewer-core (DownloadTaskEndpoint, CampaignEndpoint, AutocompleteEndpoint, UserEndpoint) accept connections without verifying that the associated HttpSession belongs to a logged-in user, and none validate the Origin header against an allowlist. An unauthenticated attacker can therefore connect to these endpoints; through CampaignEndpoint this allows polluting the in-process crowdsourcing page-lock map and broadcasting forged lock messages, disrupting reviewers' work. This issue has been patched in version 26.05.6h CVE-2026-936398.7 HIG

—

———A missing authorization vulnerability in Zimbra Collaboration Suite (ZCS) allows a delegated administrator with no explicit administrative grants to overwrite global Zimlet configuration through ConfigureZimletRequest. The uploaded configuration can also add an attacker-controlled domain to the proxy allowlist of the default class of service. If a signed-in user in that class opens a crafted /service/proxy URL, Zimbra serves attacker-controlled HTML under its own origin. The resulting JavaScript can read mailbox data and perform authenticated actions as the victim. Exploitation requires a delegated-administrator account and access to the admin SOAP service; the browser compromise additionally requires the victim to open the crafted URL. ZCS 10.1.16 and 10.1.20 are affected.5h CVE-2026-114356—

—

———Missing authorization in the system-privileged 'DJIUpdateService' binder in DJI CrystalSky ZS600C firmware v0.3.9.7 through v0.6.6.9 on Android allows a local application with no permissions to invoke privileged operations, including factory reset with user-data wipe (transaction 153), shutdown (105), clock/timezone change (101/123), and the firmware-apply hand-off (151), because onTransact() performs no caller-identity or permission check.6h CVE-2026-114354—

—

———Missing authorization in the native binder service 'android.dji.server' (transaction code 1018) in DJI CrystalSky ZS600C firmware v0.3.9.7 through v0.6.6.9 on Android allows a local application with no permissions to execute arbitrary OS commands as root, because the handler passes a caller-supplied string to system() in the root-privileged 'assistant' daemon without any caller-identity or permission check.6h CVE-2026-571169.8 CRI

—

———PraisonAI is a multi-agent teams system. From version 4.2.1 to before version 4.6.59, PraisonAI's AgentOS FastAPI deployment surface remains unauthenticated in current main and in releases after the published patched version for GHSA-pm96-6xpr-978x / CVE-2026-40151. The chat route directly calls agent.chat(request.message). No-auth and wrong-bearer requests both execute the deployed agent. This is broader than passive metadata disclosure. In any deployment where AgentOS wraps agents with tools, private context, memory, API integrations, or cost-bearing model calls, an unauthenticated reachable client can drive those agents. This issue has been patched in version 4.6.59.9h CVE-2026-568368.2 HIG

—

———PraisonAI is a multi-agent teams system. From version 4.5.112 to before version 4.6.59, PraisonAI's installed console entrypoint is Typer-first. In current releases, the recipe command is registered in the Typer app and praisonai recipe serve dispatches to the deprecated Typer command in src/praisonai/praisonai/cli/commands/recipe.py. That Typer command can start the Recipe HTTP server on a non-localhost interface with no authentication. It prints a deprecation warning, and then launches the server. Because config.auth is absent, create_app() does not attach the API-key or JWT middleware. Unauthenticated requests can then reach the recipe API and, when enabled, /admin/reload. This issue has been patched in version 4.6.59.10h