Safe2Harm: Semantic Isomorphism Attacks for Jailbreaking Large Language Models

Large Language Models LLMs have demonstrated exceptional performance across various tasks, but their security vulnerabilities can be exploited by attackers to generate harmful content, causing adverse impacts across various societal domains. Most existing jailbreak methods revolve around Prompt...

EUVD-2025-201179

An issue was discovered on Thermo Fisher Ion Torrent OneTouch 2 INS1005527 devices. When they are powered on, an X11 display server is started. The display server listens on all network interfaces and is accessible over port 6000. The X11 access control list, by default, allows connections from...

CVE-2025-11787

Command injection vulnerability in the operating system in Circutor SGE-PLC1000/SGE-PLC50 v9.0.2 through the 'GetDNS', 'CheckPing' and 'TraceRoute' functions...

CISA, Australia, and Partners Author Joint Guidance on Securely Integrating Artificial Intelligence in Operational Technology

CISA and the Australian Signals Directorate’s Australian Cyber Security Centre, in collaboration with federal and international partners, have released new cybersecurity guidance: Principles for the Secure Integration of Artificial Intelligence in Operational Technology. This guidance aims to hel...

Chopping AI Down to Size: Turning Disruptive Technology into a Strategic Advantage

Most people know the story of Paul Bunyan. A giant lumberjack, a trusted axe, and a challenge from a machine that promised to outpace him. Paul doubled down on his old way of working, swung harder, and still lost by a quarter inch. His mistake was not losing the contest. His mistake was assuming...

Picklescan Bugs Allow Malicious PyTorch Models to Evade Scans and Execute Code

Three critical security flaws have been disclosed in an open-source utility called Picklescan that could allow malicious actors to execute arbitrary code by loading untrusted PyTorch models, effectively bypassing the tool's protections. Picklescan, developed and maintained by Matthieu Maitre...

Whispering poetry at AI can make it break its own rules

Most of the big AI makers don't like people using their models for unsavory activity. Ask one of the mainstream AI models how to make a bomb or create nerve gas and you'll get the standard "I don't help people do harmful things" response. That has spawned a cat-and-mouse game of people who try to...

Is Vibe Coding Safe? Benchmarking Vulnerability of Agent-Generated Code in Real-World Tasks

Vibe coding is a new programming paradigm in which human engineers instruct large language model LLM agents to complete complex coding tasks with little supervision. Although it is increasingly adopted, are vibe coding outputs really safe to deploy in production? To answer this question, we propo...

Learning the Wrong Lessons: Syntactic-Domain Spurious Correlations in Language Models

Whitepaper from researchers at MIT, Northeastern University, and Meta. For an LLM to correctly respond to an instruction it must understand both the semantics and the domain i.e., subject area of a given task-instruction pair. However, syntax can also convey implicit information Recent work shows...

LeechHijack: Covert Computational Resource Exploitation in Intelligent Agent Systems

Large Language Model LLM-based agents have demonstrated remarkable capabilities in reasoning, planning, and tool usage. The recently proposed Model Context Protocol MCP has emerged as a unifying framework for integrating external tools into agent systems, enabling a thriving open ecosystem of...

COGNITION: From Evaluation to Defense against Multimodal LLM CAPTCHA Solvers

This paper studies how multimodal large language models MLLMs undermine the security guarantees of visual CAPTCHA. We identify the attack surface where an adversary can cheaply automate CAPTCHA solving using off-the-shelf models. We evaluate 7 leading commercial and open-source MLLMs across 18...

Systems Security Foundations for Agentic Computing

This paper articulates short- and long-term research problems in AI agent security and privacy, using the lens of computer systems security. This approach examines end-to-end security properties of entire systems, rather than AI models in isolation. While we recognize that hardening a single mode...

Securing Large Language Models (LLMs) from Prompt Injection Attacks

Large Language Models LLMs are increasingly being deployed in real-world applications, but their flexibility exposes them to prompt injection attacks. These attacks leverage the model's instruction-following ability to make it perform malicious tasks. Recent work has proposed JATMO, a task-specif...

Large Language Models Cannot Reliably Detect Vulnerabilities in JavaScript: The First Systematic Benchmark and Evaluation

Researchers have proposed numerous methods to detect vulnerabilities in JavaScript, especially those assisted by Large Language Models LLMs. However, the actual capability of LLMs in JavaScript vulnerability detection remains questionable, necessitating systematic evaluation and comprehensive...

Reverse Engineering and Control-Aware Security Analysis of the ArduPilot UAV Framework

Unmanned Aerial Vehicle UAV technologies are gaining high interest for many domains, which makes UAV security of utmost importance. ArduPilot is among the most widely used open-source autopilot UAV frameworks; yet, many studies demonstrate the vulnerabilities affecting such systems. Vulnerabiliti...

Red Teaming Large Reasoning Models

Large Reasoning Models LRMs have emerged as a powerful advancement in multi-step reasoning tasks, offering enhanced transparency and logical consistency through explicit chains of thought CoT. However, these models introduce novel safety and reliability risks, such as CoT-hijacking and...

Prompt Injection Through Poetry

In a new paper, "Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in Large Language Models," researchers found that turning LLM prompts into poetry resulted in jailbreaking the models: Abstract : We present evidence that adversarial poetry functions as a universal single-turn...

Evaluating LLMs for One-Shot Patching of Real and Artificial Vulnerabilities

Automated vulnerability patching is crucial for software security, and recent advancements in Large Language Models LLMs present promising capabilities for automating this task. However, existing research has primarily assessed LLMs using publicly disclosed vulnerabilities, leaving their...

CVE-2025-66252

Infinite Loop Denial of Service via Failed File Deletion in DB Electronica Telecomunicazioni S.p.A. Mozart FM Transmitter versions 30, 50, 100, 300, 500, 1000, 2000, 3000, 3500, 6000, 7000 allows an attacker to perform Infinite loop when unlink fails in statuscontents.php causing DoS. Due to the...

CVE-2025-66251

Unauthenticated Path Traversal with Arbitrary File Deletion in DB Electronica Telecomunicazioni S.p.A. Mozart FM Transmitter versions 30, 50, 100, 300, 500, 1000, 2000, 3000, 3500, 6000, 7000 allows an attacker to perform The deletehidden parameter allows path traversal deletion of arbitrary .tgz...