New research on AI and security, selected for relevance to language models, AI agents and cybersecurity. Summaries are the authors’ or publisher’s own words, linked to the original.
Peer-reviewed
Comparative analysis of open-source locally deployed LLMs for extraction of diagnostic information from clinical CMR imaging reports
An explainable two stage framework for spatial temporal traffic accident prediction and injury severity classification
Spatiotemporal AI-based modeling of gene regulation and variant effects in the developing human brain using Seal
ArXiv versus AI slop: will capping preprint submissions work?
Overlooked brown fat in our bodies linked to wide range of health benefits
New preprints (not yet peer-reviewed)
AgentCIScope: Grounded Contextual Integrity Evaluation for LLM Agents
LLM agents are becoming increasingly capable of handling daily workspace tasks by interacting with application environments through tool calls. However, as agents read and write sensitive user data across email, documents, and other applications, a central privacy question is whether they preserve contextual integrity (CI), that is, whether the information flows they produce comply with context-specific disclosure constraints.
EvoSafeHarness: Evolving Model- and Domain-Specific Harnesses for Securing Agents
Large language model (LLM) agents can translate natural-language requests into real-world actions, creating safety risks from both indirect prompt injections and directly harmful requests. Existing system-level safety harnesses are typically designed by experts and applied across heterogeneous models and domains, despite substantial differences in model behavior, utility tolerance, and domain-specific safety requirements.
CyberArena: Benchmarking AI Agents in Live Symmetric Attack-and-Defense Combat
AI agents are demonstrating increasingly strong offensive and defensive cybersecurity capabilities in the real world. Current benchmarks mostly evaluate agents in isolation against static targets.
P2Patch: Exploit-Guided Vulnerability Repair
Large language models are increasingly deployed as autonomous agents for automated program repair, patching software vulnerabilities through multi-step reasoning and tool use. Yet these systems typically validate a patch against the same evidence used to construct it — a static alert, sanitizer trace, published exploit, or test suite.
Natural Code Transformation Confuses Agents in Vulnerability Analysis
Having demonstrated a strong potential for automating vulnerability discovery, AI agents can also be used to attack software systems at scale. Today's software systems need mechanisms to level the playing field by slowing down potential attacks, giving them time to patch vulnerabilities and strengthen software.
Cooking Up Risks: Benchmarking and Reducing Food Safety Risks in Large Language Models
Food safety directly affects human health in daily life, as unsafe preparation or storage can lead to contamination and foodborne illness. PwC's 2025 consumer survey reports that 47% of respondents are comfortable using generative AI for meal planning, highlighting the need for safe food-related assistance.
Phantomware: a Data Augmentation Framework for the Detection of Emerging Malware Campaigns
Detecting unseen malware variants remains a major challenge as malware families continuously evolve. This issue is exacerbated in data-scarce scenarios such as emerging malware campaigns, where only a few samples are available to train robust detectors.
MMA-SafetyBench: A Benchmark for Multimodal Agent Safety Evaluation
Multimodal agents increasingly rely on visual interfaces to perceive state and select actions, creating an attack surface not captured by text-only safety evaluations. We study Visual Semantic Hijacking, in which deceptive but semantically meaningful interface elements, such as forged system alerts, cause an agent to depart from its authorized task.