Wednesday, November 18, 2026
Translate
PrivacySubmit newsShare your story Newsletter Advertise
Breaking
Research news

Research Radar, November 18, 2026: 10 new AI and security papers to know

The most relevant new AI and security research from Nature & arXiv, with the authors' own summaries and links to the papers.

10 papers, Nature & arXiv

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

Are you a budding science sleuth? Take our plagiarism quiz to test your skills

Nature

Transforming molecular evolutionary analyses with protein foundation models

Nature

The TechUpscale Brief. The day's cyber, AI and tech news in one short email, every weekday morning. Free. Unsubscribe anytime.

New preprints (not yet peer-reviewed)

Dressed in Authority: Practical Black-Box Jailbreaking of MLLMs via Visual Trust Bias

arXiv · Weihang Liu, Qianqian Xu, Yangbangyan Jiang, Zitai Wang, Zhipeng Yu, Qingming Huang

Multimodal large language models (MLLMs) are rapidly advancing, yet their safety remains a critical concern, especially under jailbreak attempts. Existing MLLM jailbreaks either inject fragile adversarial perturbations into irrelevant images or synthesize toxic visual cues to achieve text-driven attack intents.

Natural Code Transformation Confuses Agents in Vulnerability Analysis

arXiv · Jie Zhu, Ronik Bhaskar, Heather Zheng, Ben Y. Zhao, Kexin Pei

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.

SEAM: Computer-Use Agents Know the Harm Between Screens, and Commit It Anyway

arXiv · Yuxuan Zhou, Chen Xiong, Haitong Liu, Kuofeng Gao, Yang Bai, Zhiyuan He, Tao Dai, Shu-Tao Xia

Safety benchmarks for computer-use agents (CUAs) place the harm either in the intent of the task or on a single screen: a malicious instruction, an injected command, a phishing page. An agent that checks the task and each screen in isolation can pass them, so they do not test whether it combines facts seen on different screens.

GapLeak: When Privacy-Preserving LLM-based Web Agents Become a Privacy Risk

arXiv · Chaklam Cheong, Tiancong Cao, Qi Li, Zelin Zhang, Jianbing Ni, Rongxing Lu, Mohammad Zulkernine

Large Language Model (LLM)-based web agents rely on data minimization and least privilege to protect user privacy. We show that these principles are necessary but insufficient: they constrain authorized sources, but not higher-sensitivity sources that remain reachable.

Seeing with You: Perception-Reasoning Coevolution for Multimodal Reasoning

arXiv · Ziqi Miao, Haonan Jia, Lijun Li, Chen Qian, Yuan Xiong, Wenting Yan, Yongyao Wang, Zaoying Huang, Ken Deng, Jing Shao

Reinforcement learning with verifiable rewards (RLVR) has substantially enhanced the reasoning capabilities of multimodal large language models (MLLMs). However, existing RLVR approaches typically rely on outcome-driven optimization that updates both perception and reasoning using a shared reward based solely on the final answer.

WebAgents: Unbundling Agent Creation, Hosting, and Execution for an Agent-Native Web

arXiv · Dario Fenoglio, Martin Gjoreski, Marc Langheinrich, Pradyumna Chari, Ramesh Raskar

Large language models are moving from isolated assistants toward autonomous agents that can plan, use tools, and coordinate with others, motivating an emerging Agentic Web of independently developed agents and services. Realizing this vision requires agent behavior to move across creators, hosts, and runtimes without coupling publication to a particular execution environment or transferring authority over user resources.

What Does an Agent Need? Bundle-Aware Agent Resource Composition

arXiv · Yunxiao Shi (Australian National University), Tingwei Chen (University of Technology Sydney), Dannong Xu (Australian National University), Ahad N. Zehmakan (Australian National University), Jialie Shen (City St George's, University of London), Zhendong Chu (Squirrel Ai Learning), Qingsong Wen (Squirrel Ai Learning)

An agent resource's value depends on its backbone and companion resources and is known only after execution. Token-based generators remain tied to their training catalog, while independent resource scoring misses complementarities and reused bundles target other tasks.

From Strategy Search to Policy Learning: MCTS-Guided Adversarial Co-Evolution for LLM Safety Alignment

arXiv · Xurui Li, Kaisong Song

Large Language Models (LLMs) increasingly power web-facing applications, including conversational assistants and content services, where open user interaction exposes them to diverse safety threats and adaptive attacks. In particular, jailbreak attackers repeatedly reformulate harmful requests to bypass safety alignment.

Add TechUpscale as a preferred source on Google

The TechUpscale Brief

The day's cyber, AI and tech news in one short email, every weekday morning. Free. Unsubscribe anytime.

I'm most interested in

Join the conversation

Comments are moderated. Be civil and stay on the story: no personal attacks, hate, spam or ads, and no one else's private details. Your email is never shown. Privacy policy.

More Research