Results for “Again”

arXiv (research papers) Saeid Jamshidi SI

Semantic Attacks on Tool-Augmented LLMs: Securing the Model Context Protocol Against Descriptor-Level Manipulation

The Model Context Protocol (MCP) enables Large Language Models (LLMs) to interact with external tools via tool descriptors, thereby extending their capabilities for task execution, autonomous decision-making, and multi-agent coordination. Existing MCP deployments treat tool descriptors as trusted metadata, despite their direct integration into the LLM reasoning context. This introduces a previously underexplored semantic attack surface. Current defenses primarily target…

Read at arXiv (research papers)

arXiv (research papers) Weikai Lu SI

ARGUS: Defending Against Multimodal Indirect Prompt Injection via Steering Instruction-Following Behavior

Multimodal Large Language Models (MLLMs) are increasingly vulnerable to multimodal Indirect Prompt Injection (IPI) attacks, which embed malicious instructions in images, videos, or audio to hijack model behavior. Existing defenses, designed primarily for text-only LLMs, are unsuitable for countering these multimodal threats, as they are easily bypassed, modality-dependent, or generalize poorly. Inspired by activation steering researches, we hypothesize that a robust, general…

Read at arXiv (research papers)

Journal articles (OpenAlex) Mohammad Amaan Siddiqui · Nations and Nationalism Geopolitics

Rhetorics of Counternationalism: The Limitations of Digital Anti‐Hindutva in Combating Right‐Wing Extremism

ABSTRACT How are online discourses in subissues within counternationalist movements constructed? This study better understands what comprises digital counternationalist dissent against right‐wing nationalism, finding that right‐wing nationalism's success can also be explained through limitations in counternationalist discourse. In this project, three coders analysed dissent against India's right‐wing nationalism, Hindutva, parsing anti‐Hindutva into different themes for…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Jaime Cubides Cárdenas, Christian Esteban Hernández-Alvarado, Jonnathan Jiménez Reina et al. · Frontiers in Political Science Geopolitics

Regional cooperation against transnational organized crime: challenges and opportunities in the Western Hemisphere

Transnational Organized Crime represents a critical threat to the political, economic, and social stability of the Western Hemisphere. This article analyses the dynamics of the criminal organizations that constitute this phenomenon, including their illicit activities such as drug, human, and arms trafficking, as well as their relationship with migratory flows and advanced technologies. Using a qualitative methodology based on the multidimensional security model, it evaluates…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Guilherme Cavalcante Silva · Tapuya Latin American Science Technology and Society Geopolitics

Regional integration as a remedy against technological dependency? The contradictions in the calls for regional integration in Latin American AI strategies

The rise of regional integration and South-South cooperation as an imperative in international relations and trade has been widely discussed, especially with the strengthening of BRICS+ (Brazil, Russia, India, China, South Africa), and Pan-African movements. The idea of a world with multiple powers giving rise to diverse social and economic imperatives for technoscientific innovation is finding traction among scholars, including researchers in science and technology studies…

Read at Journal articles (OpenAlex)

Naval News Aaron-Matthew Lariosa South China Sea

Philippines Challenges Chinese Forces at Scarborough Shoal

Two Philippine maritime patrols challenged the presence of China’s warships and cutters at Scarborough Shoal over the last few days, marking Manila’s latest effort to push back against Beijing’s claims over the South China Sea maritime feature. Last Friday saw a Bureau of Fisheries and Aquatic Resources maritime domain awareness flight receive up to 40 ...

Read at Naval News

arXiv (research papers) Yanxi Li SI

Semantics as a Shield: Label Disguise Defense (LDD) against Prompt Injection in LLM Sentiment Classification

Prompt injection can cause large language models (LLMs) to deviate from their intended tasks by following instructions embedded in untrusted input. We investigate whether changing a classifier's output labels improves accuracy under such attacks, using sentiment classification as a case study. Label Disguise Defense (LDD) replaces positive/negative with alternatives such as good/bad or cat/dog, introduced through demonstrations, and maps predictions back to sentiment…

Read at arXiv (research papers)