CVE-2026-57474

The CVE entry CVE-2026-57474 concerns Deloitte AI Assist for Customer. Public-facing API endpoints accepted unauthenticated requests, exposing configuration information that could aid an attacker’s reconnaissance. The root cause is exposure of configuration data via unauthenticated access to API ...

CVE-2026-1946

The GW AI Website Builder plugin for WordPress is vulnerable to unauthorized modification of data due to a missing capability check on the gwaiwebugravitywritedisconnecthandler function in all versions up to, and including, 1.0.1. This makes it possible for authenticated attackers, with...

Improper Certificate Validation

Coder is vulnerable to Improper Certificate Validation. The vulnerability is due to the AI Bridge Proxy aibridgeproxyd creating a default HTTPS transport with InsecureSkipVerify enabled when no upstream proxy is configured, which allows an on-path man-in-the-middle attacker to present a forged TL...

CVE-2026-1946

The GW AI Website Builder plugin for WordPress is vulnerable to unauthorized modification of data due to a missing capability check on the gwaiwebugravitywritedisconnecthandler function in all versions up to, and including, 1.0.1. This makes it possible for authenticated attackers, with...

CVE-2026-15286

The Gutenberg Blocks with AI by Kadence WP – Page Builder Features plugin for WordPress is vulnerable to unauthorized post publication in all versions up to, and including, 3.5.32 due to a misconfigured capability check on the 'getitemspermissioncheck' function permission callback of the...

EUVD-2026-42795

The Gutenberg Blocks with AI by Kadence WP – Page Builder Features plugin for WordPress is vulnerable to unauthorized post publication in all versions up to, and including, 3.5.32 due to a misconfigured capability check on the 'getitemspermissioncheck' function permission callback of the...

CVE-2026-15286 Gutenberg Blocks with AI by Kadence WP – Page Builder Features <= 3.5.32 - Incorrect Authorization to Authenticated (Contributor+) Post Publication

The Gutenberg Blocks with AI by Kadence WP – Page Builder Features plugin for WordPress is vulnerable to unauthorized post publication in all versions up to, and including, 3.5.32 due to a misconfigured capability check on the 'getitemspermissioncheck' function permission callback of the...

Coder 2.30.x < 2.32.7 / 2.33.x < 2.33.8 / 2.34.x < 2.34.2 AI Bridge Proxy TLS Verification Bypass

The version of Coder installed on the remote host is 2.30.x prior to 2.32.7, 2.33.x prior to 2.33.8, or 2.34.x prior to 2.34.2. It is, therefore, affected by a vulnerability. The AI Bridge Proxy skips TLS certificate verification in the default configuration. The proxy created a goproxy server...

PT-2026-57120

The GW AI Website Builder plugin for WordPress is vulnerable to unauthorized modification of data due to a missing capability check on the gwaiwebu gravitywrite disconnect handler function in all versions up to, and including, 1.0.1. This makes it possible for authenticated attackers, with...

AI Gateway Connected to Amazon Bedrock Hijacked for Cryptomining

Darktrace says a LiteLLM AI gateway linked to Amazon Bedrock was compromised for cryptomining after signs of exposed SSH activity...

CVE-2026-59207

n8n is an open source workflow automation platform. Prior to 2.27.4 and 2.28.1, the AI Agents feature did not enforce the Allowed HTTP Request Domains restriction configured on credentials when an MCP tool was pointed at an arbitrary URL, allowing a member-level user with use-only access to a...

CVE-2026-59207 n8n: "Allowed HTTP Request Domains" Restriction Bypass via AI Agents MCP Connector

n8n is an open source workflow automation platform. Prior to 2.27.4 and 2.28.1, the AI Agents feature did not enforce the Allowed HTTP Request Domains restriction configured on credentials when an MCP tool was pointed at an arbitrary URL, allowing a member-level user with use-only access to a...

CVE-2026-59207 n8n: "Allowed HTTP Request Domains" Restriction Bypass via AI Agents MCP Connector

n8n is an open source workflow automation platform. Prior to 2.27.4 and 2.28.1, the AI Agents feature did not enforce the Allowed HTTP Request Domains restriction configured on credentials when an MCP tool was pointed at an arbitrary URL, allowing a member-level user with use-only access to a...

CVE-2026-59207

The CVE covers n8n (open source workflow automation) where, prior to versions 2.27.4 and 2.28.1, the AI Agents feature did not enforce the Allowed HTTP Request Domains restriction configured on credentials when an MCP tool targeted an arbitrary URL. This allowed a member-level user with use-only ...

Important: Red Hat Security Advisory: RHTAS 1.4.2 - Tech Preview Release Of the Go based Model Transparency CLI

The Tech Preview release of the Go based RHTAS Model Transparency CLI image. For more details please visit the product documentation at https://access.redhat.com/documentation/en-us/redhattrustedartifactsigner/1.4 The Go based RHTAS Model Transparency CLI image can be used to sign and verify AI/M...

exploitarium

https://discord.gg/WytKH65ZR join up for research, help, documen...

Meta's New AI Image Tool Lets Others Use Your Public Instagram Photos in AI Images

Meta has announced that its new artificial intelligence AI model Muse Image lets people use public Instagram posts and reels to generate AI content, and it's enabled by default. "You can also @-mention Instagram accounts in the Meta AI app to bring specific Instagram profiles right into your...

Top AI Agents Built to Catch Malicious Code Can Be Tricked Into Running It

Ask an AI coding agent to scan open-source code for security holes, and it might run the attacker's code on your own machine instead. That is the finding in a proof-of-concept published Wednesday by the AI Now Institute, an attack it calls "Friendly Fire. " It works against Anthropic's Claude Cod...

PT-2026-56832

Name of the Vulnerable Software and Affected Versions Open WebUI versions 0.6.16 through 0.9.x Description The Socket.IO server is configured with always connect=True, which allows the ydoc:awareness:update and ydoc:document:leave Socket.IO handlers to accept collaborative-document events without...

PT-2026-56885

Name of the Vulnerable Software and Affected Versions Open WebUI versions 0.7.0 through 0.9.9 Description A flaw exists where the endpoint "/api/v1/channels//members" returns full UserModelResponse objects for channel members. This allows a standard channel participant to retrieve sensitive...