The Trigger in the Haystack: Extracting and Reconstructing LLM Backdoor Triggers

Detecting whether a model has been poisoned is a longstanding problem in AI security. In this work, we present a practical scanner for identifying sleeper agent-style backdoors in causal language models. Our approach relies on two key findings: first, sleeper agents tend to memorize poisoning dat...

RPP: A Certified Poisoned-Sample Detection Framework for Backdoor Attacks under Dataset Imbalance

Deep neural networks are highly susceptible to backdoor attacks, yet most defense methods to date rely on balanced data, overlooking the pervasive class imbalance in real-world scenarios that can amplify backdoor threats. This paper presents the first in-depth investigation of how the dataset...

Mustang Panda Deploys Updated COOLCLIENT Backdoor in Government Cyber Attacks

Threat actors with ties to China have been observed using an updated version of a backdoor called COOLCLIENT in cyber espionage attacks in 2025 to facilitate comprehensive data theft from infected endpoints. The activity has been attributed to Mustang Panda aka Earth Preta, Fireant, HoneyMyte,...

Exploit for Server-Side Request Forgery in Microsoft

CTT-ProxyLogon-RCE-v1.0---Convergent-Time-Theory-Enhanced-Micr...

Konni Hackers Deploy AI-Generated PowerShell Backdoor Against Blockchain Developers

The North Korean threat actor known as Konni has been observed using PowerShell malware generated using artificial intelligence AI tools to target developers and engineering teams in the blockchain sector. The phishing campaign has targeted Japan, Australia, and India, highlighting the adversary'...

Exploit for Improper Input Validation in N8N

🛡️ RulesGuard !Testshttps://github.com/NOTTIBOY137/RulesG...

Phishing Attack Uses Stolen Credentials to Install LogMeIn RMM for Persistent Access

Cybersecurity researchers have disclosed details of a new dual-vector campaign that leverages stolen credentials to deploy legitimate Remote Monitoring and Management RMM software for persistent remote access to compromised hosts. "Instead of deploying custom viruses, attackers are bypassing...

CVE-2026-1331

MeetingHub developed by HAMASTAR Technology has an Arbitrary File Upload vulnerability, allowing unauthenticated remote attackers to upload and execute web shell backdoors, thereby enabling arbitrary code execution on the server...

CVE-2025-56005

An arbitrary code execution vulnerability was discovered in PLY Python Lex-Yacc. When an application uses PLY's undocumented picklefile parameter to load cached parser data, the library deserializes the pickle file without validation. If an attacker can supply or modify the pickle file being...

SUSE CVE-2025-56005

An undocumented and unsafe feature in the PLY Python Lex-Yacc library 3.11 allows Remote Code Execution RCE via the picklefile parameter in the yacc function. This parameter accepts a .pkl file that is deserialized with pickle.load without validation. Because pickle allows execution of embedded...

HAMASTAR MeetingHub code-related vulnerabilities

HAMASTAR MeetingHub is a paperless conference system developed by HAMASTAR, a company from Taiwan, China. HAMASTAR MeetingHub has code-related vulnerabilities. These vulnerabilities stem from an arbitrary file upload vulnerability, which may allow unverified remote attackers to upload and execute...

20,000 WordPress Sites Affected by Backdoor Vulnerability in LA-Studio Element Kit for Elementor WordPress Plugin

On January 12th, 2026, we received a submission for a Backdoor vulnerability in the LA-Studio Element Kit for Elementor, a WordPress plugin with more than 20,000+ active installations. This vulnerability makes it possible for an unauthenticated attacker to create malicious administrator users...

CVE-2026-1222

PrismX MX100 AP controller developed by BROWAN COMMUNICATIONS has an Arbitrary File Upload vulnerability, allowing privileged remote attackers to upload and execute web shell backdoors, thereby enabling arbitrary code execution on the server...

CVE-2025-56005

An undocumented and unsafe feature in the PLY Python Lex-Yacc library 3.11 allows Remote Code Execution RCE via the picklefile parameter in the yacc function. This parameter accepts a .pkl file that is deserialized with pickle.load without validation. Because pickle allows execution of embedded...

North Korea-Linked Hackers Target Developers via Malicious VS Code Projects

The North Korean threat actors associated with the long-running Contagious Interview campaign have been observed using malicious Microsoft Visual Studio Code VS Code projects as lures to deliver a backdoor on compromised endpoints. The latest finding demonstrates continued evolution of the new...

CVE-2026-1222 BROWAN COMMUNICATIONS ｜PrismX MX100 AP controller - Arbitrary File Upload

PrismX MX100 AP controller developed by BROWAN COMMUNICATIONS has an Arbitrary File Upload vulnerability, allowing privileged remote attackers to upload and execute web shell backdoors, thereby enabling arbitrary code execution on the server...

CVE-2025-56005

An undocumented and unsafe feature in the PLY Python Lex-Yacc library 3.11 allows Remote Code Execution RCE via the picklefile parameter in the yacc function. This parameter accepts a .pkl file that is deserialized with pickle.load without validation. Because pickle allows execution of embedded...

CVE-2025-56005

An undocumented and unsafe feature in the PLY Python Lex-Yacc library 3.11 allows Remote Code Execution RCE via the picklefile parameter in the yacc function. This parameter accepts a .pkl file that is deserialized with pickle.load without validation. Because pickle allows execution of embedded...

A Survey of Security Challenges and Solutions for Advanced Air Mobility and EVTOL Aircraft

This survey reviews the existing and envisioned security vulnerabilities and defense mechanisms relevant to Advanced Air Mobility AAM systems, with a focus on electric vertical takeoff and landing eVTOL aircraft. Drawing from vulnerabilities in the avionics in commercial aviation and the automate...

SecureSplit: Mitigating Backdoor Attacks in Split Learning

Split Learning SL offers a framework for collaborative model training that respects data privacy by allowing participants to share the same dataset while maintaining distinct feature sets. However, SL is susceptible to backdoor attacks, in which malicious clients subtly alter their embeddings to...