Related papers: STACK: Adversarial Attacks on LLM Safeguard Pipeli…

Many cache designs have been proposed to guard against contention-based side-channel attacks. One well-known type of cache is the randomized remapping cache. Many randomized remapping caches provide fixed or over protection, which leads to…

Deep research frameworks have shown promising capabilities in synthesizing comprehensive reports from web sources. While deep research possesses significant potential to address complex issues through planning and research cycles, existing…

Foundation models represent the most prominent and recent paradigm shift in artificial intelligence. Foundation models are large models, trained on broad data that deliver high accuracy in many downstream tasks, often without fine-tuning.…

As frontier AI models become more capable, evaluating their potential to enable cyberattacks is crucial for ensuring the safe development of Artificial General Intelligence (AGI). Current cyber evaluation efforts are often ad-hoc, lacking…

Safety, security, and compliance are essential requirements when aligning large language models (LLMs). However, many seemingly aligned LLMs are soon shown to be susceptible to jailbreak attacks. These attacks aim to circumvent the models'…

As large language models (LLMs) increasingly integrate native code interpreters, they enable powerful real-time execution capabilities, substantially expanding their utility. However, such integrations introduce potential system-level…

The transition of Large Language Models (LLMs) from passive code generators to autonomous agents introduces significant safety risks, specifically regarding destructive commands and inconsistent system states. Existing commercial solutions…

In this study, a modular, data-free pipeline for multi-label intention recognition is proposed for agentic AI applications in transportation. Unlike traditional intent recognition systems that depend on large, annotated corpora and often…

Autonomous AI agents are being deployed with filesystem access, email control, and multi-step planning. This thesis contributes to four open problems in AI safety: understanding dangerous internal computations, removing dangerous behaviors…

Large Language Models (LLMs) deploy safety mechanisms to prevent harmful outputs, yet these defenses remain vulnerable to adversarial prompts. While existing research demonstrates that jailbreak attacks succeed, it does not explain…

Artificial Intelligence's dual-use nature is revolutionizing the cybersecurity landscape, introducing new threats across four main categories: deepfakes and synthetic media, adversarial AI attacks, automated malware, and AI-powered social…

An agentic-AI runtime issues tool calls, sends messages, and actuates devices on behalf of an LLM. Catching the four ways an action can diverge from its audit record -- F1 gate-bypass, F2 audit-forgery, silent host failure, F4 wrong-target,…

Powerful autonomous systems, which reason, plan, and converse using and between numerous tools and agents, are made possible by Large Language Models (LLMs), Vision-Language Models (VLMs), and new agentic AI systems, like LangChain and…

While large language models (LLMs) have achieved remarkable success in providing trustworthy responses for knowledge-intensive tasks, they still face critical limitations such as hallucinations and outdated knowledge. To address these…

New hardware primitives such as Intel SGX secure a user-level process in presence of an untrusted or compromised OS. Such "enclaved execution" systems are vulnerable to several side-channels, one of which is the page fault channel. In this…

Per Row Activation Counting (PRAC) has emerged as a robust framework for mitigating RowHammer (RH) vulnerabilities in modern DRAM systems. However, we uncover a critical vulnerability: a timing channel introduced by the Alert Back-Off (ABO)…

This work addresses the challenges related to attacks on collaborative tagging systems, which often comes in a form of malicious annotations or profile injection attacks. In particular, we study various countermeasures against two types of…

Transparent authentication (TA) schemes are those in which a user is authenticated by a verifier without requiring explicit user interaction. By doing so, those schemes promise high usability and security simultaneously. The majority of TA…

Prompt injection remains one of the most practical attack vectors against LLM-integrated applications. We replicate the Microsoft LLMail-Inject benchmark (Greshake et al., 2024) against current generation models running inside OpenClaw, an…

LLM-powered applications routinely embed secrets in system prompts, yet models can be tricked into revealing them. We built an adaptive attacker that evolves its strategies over hundreds of rounds and tested it against nine defense…