Updating the taxonomy of failure modes in agentic AI systems: What a year of red teaming taught us

In this article 1. Why the Taxonomy Needed Updating 2. Seven new failure modes 3. Operational findings: What red teaming showed 4. New mitigations 5. What to do this quarter When the Microsoft AI Red Team published the Taxonomy of Failure Modes in Agentic AI Systems in April 2025, the goal was a...

Imperva Customers Protected Against CVE-2026-49975 (HTTP/2 Bomb) DoS

TL;DR: CVE-2026-49975, dubbed the “HTTP/2 Bomb,” is a critical remote Denial-of-Service DoS vulnerability affecting default HTTP/2 configurations of major web servers including NGINX, Apache HTTPD, Microsoft IIS, Envoy, and Cloudflare Pingora. Discovered by security firm Calif using OpenAI’s Code...

GHSA-P92Q-9VQR-4J8V Axios: Proxy-Authorization Credential Leak to Origin Server Across HTTP-to-HTTPS Redirect in Axios Node.js HTTP Adapter

Summary Axios’s Node.js HTTP adapter may forward a Proxy-Authorization header to a redirected origin during specific proxy-to-direct redirect flows. This affects Node.js usage, where an initial HTTP request is sent through an authenticated HTTP proxy, redirects are followed, and the redirected UR...

Exploit for CVE-2026-49975

CVE-2026-49975 — HTTP/2 Bomb PoC !CVEhttps://img.shields...

Ubuntu 22.04 LTS / 24.04 LTS / 25.10 / 26.04 LTS : FRR vulnerabilities (USN-8376-1)

The remote Ubuntu 22.04 LTS / 24.04 LTS / 25.10 / 26.04 LTS host has packages installed that are affected by multiple vulnerabilities as referenced in the USN-8376-1 advisory. It was discovered that FRR incorrectly handled certain OSPF Traffic Engineering and Segment Routing TLVs. An attacker cou...

PT-2026-46850

Summary There is a Proof of Concept which is able to enumerate the usernames of administrator users. This was possible by performing a timing attack. Details The faulty code exists in src/Core/Framework/Api/OAuth/UserRepository.php: public function getUserEntityByUserCredentials string $username,...

CVE-2026-42849

authentik is an open-source identity provider. Prior to versions 2025.12.5 and 2026.2.3, due to the implementation of stages in the SFE Simple Flow Executor in order to make the interface more compatible with legacy browsers, it was possible to use an XSS exploit in the AutosubmitStage. This issu...

auto-survey (>=0.1.0 <=0.2.5), gptparse (=0.3.0) +12 more potentially affected by CVE-2026-44018 via docling (>=2.51.0 <=2.90.0)

docling PYPI version =2.51.0, =0.1.0, =0.19.2, =1.40.0, =0.6.2, =0.0.1, =0.3.0, =1.0.0, =1.6.2, =1.6.2, =0.2.4, =0.0.1, =0.0.2 Source cves: CVE-2026-44018 Source advisory: OSV:GHSA-R3XG-RG9J-67FV...

auto-survey (>=0.1.0 <=0.2.5), gptparse (=0.3.0) +5 more potentially affected by CVE-2026-44016 via docling (>=2.87.0 <=2.90.0)

docling PYPI version =2.87.0, =0.1.0, =0.40.0, =0.6.2, =0.0.1, =0.0.1, =0.0.2 Source cves: CVE-2026-44016 Source advisory: SNYK:PYTHON-DOCLING-17151857...

CVE-2026-46266

A flaw was found in the Linux kernel's handling of RAW sockets using IPPROTORAW. A remote attacker could send a specially crafted ICMP Internet Control Message Protocol packet. This malicious packet could set the protocol field to 255, causing it to be processed by a RAW socket configured for...

USN-8376-1: FRR vulnerabilities

It was discovered that FRR incorrectly handled certain OSPF Traffic Engineering and Segment Routing TLVs. An attacker could possibly use this issue to cause FRR to crash, resulting in a denial of service. CVE-2026-28532 It was discovered that FRR incorrectly handled certain BGP FlowSpec component...

USN-8376-1 frr vulnerabilities

It was discovered that FRR incorrectly handled certain OSPF Traffic Engineering and Segment Routing TLVs. An attacker could possibly use this issue to cause FRR to crash, resulting in a denial of service. CVE-2026-28532 It was discovered that FRR incorrectly handled certain BGP FlowSpec component...

CVE-2026-24090

Cryptographic issue while processing partition table entries allows unauthorized modification of boot flow...

Security Bulletin: IBM Maximo Application Suite uses once-2.0.0.tgz which is vulnerable to CVE-2026-3449

Summary IBM Maximo Application Suite uses once-2.0.0.tgz which is vulnerable to CVE-2026-3449. This bulletin contains information regarding the vulnerability and its fixture. Vulnerability Details CVEID:CVE-2026-3449 DESCRIPTION: Versions of the package @tootallnate/once before 3.0.1 are vulnerab...

PT-2026-46006

Name of the Vulnerable Software and Affected Versions OP-TEE versions 3.16.0 through 4.10.x Description A use-after-free race condition exists in the shared memory teardown logic of FF-A within SPMC/SP flows. This occurs when OP-TEE is configured as an SPMC for S-EL0 SPs using CFG SECURE...

📄 Gogs Git Rebase Argument Injection / Remote Code Execution

This Metasploit module exploits an argument injection vulnerability in the pull request merge flow of Gogs versions less than or equal to 0.14.2 and less than or equal to 0.15.0+dev. frozenstringliteral: true This module requires Metasploit: https://metasploit.com/download Current source:...

CVE-2026-42849

authentik is an open-source identity provider. Prior to versions 2025.12.5 and 2026.2.3, due to the implementation of stages in the SFE Simple Flow Executor in order to make the interface more compatible with legacy browsers, it was possible to use an XSS exploit in the AutosubmitStage. This issu...

EUVD-2026-34026

authentik is an open-source identity provider. Prior to versions 2025.12.5 and 2026.2.3, due to the implementation of stages in the SFE Simple Flow Executor in order to make the interface more compatible with legacy browsers, it was possible to use an XSS exploit in the AutosubmitStage. This issu...

CVE-2026-42073

OpenClaude is an open-source coding-agent command line interface for cloud and local model providers. Prior to version 0.5.1, the OpenClaude MCP authentication flow starts a temporary local HTTP server to handle OAuth callbacks. To prevent CSRF attacks, the server validates a state parameter...

CVE-2026-48134

When the DLP is active, the UserCheck Web Portal contains an input-handling issue in the UserChoice flow. Under specific conditions, an attacker who can access the UserCheck Ask page could attempt to manipulate the Security Gateway's stored DLP/UserCheck incident information. This could lead to...