Bypassing AI Control Protocols Via Agent-As-A-Proxy Attacks

As AI agents automate critical workloads, they remain vulnerable to indirect prompt injection IPI attacks. Current defenses rely on monitoring protocols that jointly evaluate an agent's Chain-of-Thought CoT and tool-use actions to ensure alignment with user intent. We demonstrate that these...

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...

Stealthy Poisoning Attacks Bypass Defenses in Regression Settings

Regression models are widely used in industrial processes, engineering and in natural and physical sciences, yet their robustness to poisoning has received less attention. When it has, studies often assume unrealistic threat models and are thus less useful in practice. In this paper, we propose a...

A Systematic Literature Review on LLM Defenses against Prompt Injection and Jailbreaking: Expanding NIST Taxonomy

The rapid advancement and widespread adoption of generative artificial intelligence GenAI and large language models LLMs has been accompanied by the emergence of new security vulnerabilities and challenges, such as jailbreaking and other prompt injection attacks. These maliciously crafted inputs...

Johnson Controls Inc. iSTAR Configuration Utility (ICU) tool

RISK EVALUATION Successful exploitation of this vulnerability could allow an attacker to cause a failure within the operating system of the machine hosting the ICU tool. 2. RECOMMENDED PRACTICES CISA recommends users take defensive measures to minimize the risk of exploitation of this...

Top 10 Threat Actors: Their Tactics & Motivations

The world of cybercrime has its own cast of characters, each with a unique script. Some are patient spies, like state-sponsored groups that move silently within a network for months to gather intelligence. Others are loud and aggressive, like ransomware gangs that operate like ruthless businesses...

Sockpuppetting: Jailbreaking LLMs without Optimization through Output Prefix Injection

As open-weight large language models LLMs increase in capabilities, safeguarding them against malicious prompts and understanding possible attack vectors becomes ever more important. While automated jailbreaking methods like GCG Zou et al., 2023 remain effective, they often require substantial...

PT-2026-1688

Name of the Vulnerable Software and Affected Versions affected versions not specified Description Attackers are leveraging artificial intelligence to customize malicious payloads in phishing campaigns, enabling them to circumvent security defenses. This issue is related to email security protocol...

Jailbreaking Large Language Models through Iterative Tool-Disguised Attacks Via Reinforcement Learning

Large language models LLMs have demonstrated remarkable capabilities across diverse applications, however, they remain critically vulnerable to jailbreak attacks that elicit harmful responses violating human values and safety guidelines. Despite extensive research on defense mechanisms, existing...

Research Directions in Quantum Computer Cybersecurity

This document presents a concise overview of the contemporary research directions in quantum computer cybersecurity. The aim of this document is not to be a survey, but rather a succinct summary of the major research directions in quantum computer cybersecurity at the end of the first half of the...

Cryptanalysis of Pseudorandom Error-Correcting Codes

Pseudorandom error-correcting codes PRC is a novel cryptographic primitive proposed at CRYPTO 2024. Due to the dual capability of pseudorandomness and error correction, PRC has been recognized as a promising foundational component for watermarking AI-generated content. However, the security of PR...

PT-2025-52309

3/ CVE-2023-1423 highlights vulnerabilities exploited by AI-driven phishing, emphasizing the need for updated defenses. CyberSecurity Hacking...

CIS-BA: Continuous Interaction Space Based Backdoor Attack for Object Detection in the Real-World

Object detection models deployed in real-world applications such as autonomous driving face serious threats from backdoor attacks. Despite their practical effectiveness,existing methods are inherently limited in both capability and robustness due to their dependence on single-trigger-single-objec...

Turn me on, turn me off: Zigbee assessment in industrial environments

We all encounter IoT and home automation in some form or another, from smart speakers to automated sensors that control water pumps. These services appear simple and straightforward to us, but many devices and protocols work together under the hood to deliver them. One of those protocols is Zigbe...

Google Adds Layered Defenses to Chrome to Block Indirect Prompt Injection Threats

Google on Monday announced a set of new security features in Chrome, following the company's addition of agentic artificial intelligence AI capabilities to the web browser. To that end, the tech giant said it has implemented layered defenses to make it harder for bad actors to exploit indirect...

Chrome Zero-Day Vulnerability: Risks & Protection

Your team knows the drill: a security alert goes out, and everyone scrambles to patch. But what happens in the critical window before a fix is available for a new Chrome zero-day vulnerability? Relying on a reactive cycle of patching leaves your organization dangerously exposed. Attackers thrive ...

Inside Hive Pro: A Complete Platform Review

Knowing you have a vulnerability is one thing; knowing if you’re truly exposed is another. A critical vulnerability might exist on a server, but can an attacker actually reach it? Will your firewall block the attempt? Will your EDR detect the payload? Traditional vulnerability management can't...

Defending Large Language Models against Jailbreak Exploits with Responsible AI Considerations

Large Language Models LLMs remain susceptible to jailbreak exploits that bypass safety filters and induce harmful or unethical behavior. This work presents a systematic taxonomy of existing jailbreak defenses across prompt-level, model-level, and training-time interventions, followed by three...

Breaking Down S3 Ransomware: Variants, Attack Paths and Trend Vision One™ Defenses

In this blog entry, Trend™ Research explores how ransomware actors are shifting their focus to cloud-based assets, including the tactics used to compromise business-critical data in AWS environments...

Jailbreaking Large Vision Language Models in Intelligent Transportation Systems

Large Vision Language Models LVLMs demonstrate strong capabilities in multimodal reasoning and many real-world applications, such as visual question answering. However, LVLMs are highly vulnerable to jailbreaking attacks. This paper systematically analyzes the vulnerabilities of LVLMs integrated ...