Results for “face”

arXiv (research papers) Yanting Wang SI

FlashRT: Towards Computationally and Memory Efficient Red-Teaming for Prompt Injection and Knowledge Corruption

Long-context large language models (LLMs)-for example, Gemini-3.1-Pro and Qwen-3.5-are widely used to empower many real-world applications, such as retrieval-augmented generation, autonomous agents, and AI assistants. However, security remains a major concern for their widespread deployment, with threats such as prompt injection and knowledge corruption. To quantify the security risks faced by LLMs under these threats, the research community has developed heuristic-based and…

Read at arXiv (research papers)

arXiv (research papers) Zonghao Ying SI

AgentVisor: Defending LLM Agents Against Prompt Injection via Semantic Virtualization

Large Language Model (LLM) agents are increasingly used to automate complex workflows, but integrating untrusted external data with privileged execution exposes them to severe security risks, particularly direct and indirect prompt injection. Existing defenses face significant challenges in balancing security with utility, often encountering a trade-off where rigorous protection leads to over-defense, or where subtle indirect injections bypass detection. Drawing inspiration…

Read at arXiv (research papers)

arXiv (research papers) Jiaqi Li SI

Poster: ClawdGo: Endogenous Security Awareness Training for Autonomous AI Agents

Autonomous AI agents deployed on platforms such as OpenClaw face prompt injection, memory poisoning, supply-chain attacks, and social engineering, yet existing defences address only the platform perimeter, leaving the agent's own threat judgement entirely untrained. We present ClawdGo, a framework for endogenous security awareness training: we teach the agent to recognise and reason about threats from the inside, at inference time, with no model modification. Four…

Read at arXiv (research papers)

Journal articles (OpenAlex) Amanda Moeller, Samuel Hunter · Trends in Organized Crime Geopolitics

Narcos, Inc.: drug cartels and the architecture of entrepreneurial ecosystems

Abstract As the U.S. State Department increasingly designates drug cartels as Foreign Terrorist Organizations (FTOs), scholars and policymakers face a pressing need to better understand the structure and behavior of these groups. This paper argues that drug cartels are more accurately conceptualized as criminal entrepreneurial ecosystems than as traditional terrorist organizations. Drawing on the entrepreneurial ecosystem (EE) framework, which emphasizes interdependence…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Zhang Renran · Higher Education Studies Geopolitics

Generational Trauma and Identity Construction in Cristina García’s Dreaming in Cuban and The Agüero Sisters

As a core representative of Cuban-American literature, Cristina García focuses her creations on the exile experiences of Cubans after the Cuban Revolution, and profoundly depicts the identity predicament and intergenerational transmission of trauma faced by the Cuban-American “one-and-a-half generation”. Taking her representative works Dreaming in Cuban (1992) and The Agüero Sisters (1997) as the research objects, this paper systematically analyzes the specific…

Read at Journal articles (OpenAlex)

Latinoamérica21 William A. Clavijo Geopolitics

The Strait of Hormuz and the Fragility of the Global Energy System

The blockade of the Strait of Hormuz shows how a single critical chokepoint can destabilize the global energy system, exposing its profound fragility in the face of geopolitical conflicts. O post The Strait of Hormuz and the Fragility of the Global Energy System apareceu primeiro em Latinoamérica 21.

Read at Latinoamérica21

arXiv (research papers) Runpeng Geng SI

PIArena: A Platform for Prompt Injection Evaluation

Prompt injection attacks pose serious security risks across a wide range of real-world applications. While receiving increasing attention, the community faces a critical gap: the lack of a unified platform for prompt injection evaluation. This makes it challenging to reliably compare defenses, understand their true robustness under diverse attacks, or assess how well they generalize across tasks and benchmarks. For instance, many defenses initially reported as effective were…

Read at arXiv (research papers)

arXiv (research papers) Wenkui Yang SI

Are GUI Agents Focused Enough? Automated Distraction via Semantic-level UI Element Injection

Existing red-teaming studies on GUI agents face two fundamental limitations: adversarial perturbations require white-box access unavailable in commercial deployments, while prompt injection is increasingly neutralized by stronger safety alignment. To study robustness under a more practical threat model, we propose Semantic-level UI Element Injection, a black-box red-teaming paradigm that overlays safety-aligned and harmless UI elements onto screenshots to misdirect the…

Read at arXiv (research papers)

Journal articles (OpenAlex) Madhav Joshi, Grace Sullivan · Australian Journal Of International Affairs Geopolitics

Where are the international allies of Afghan women?

The Women, Peace, and Security (WPS) agenda has provided a gendered governance framework for conflict prevention and international peacebuilding for the last twenty-five years. Under the Taliban’s rule in Afghanistan, women and girls face ‘gender apartheid,’ including bans on education and employment, exclusion from public spaces, and widespread gender-based violence. This discussion presents data on the international community’s interactions with the Taliban collected…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Jhon Valdiglesias · F1000Research Geopolitics

Regulatory Design and Long-Term Growth: Insights from China’s Experience for Latin America

Background: Regulatory efficiency is increasingly recognized as a key determinant of long-term economic growth. China's institutional consolidation offers a comparative reference case for understanding how streamlined procedures, and coordinated governance can reduce transaction costs and foster development. In contrast, many Latin American countries face persistent administrative complexity that may hinder growth. Methods: This study applies Institutional Theory and…

Read at Journal articles (OpenAlex)

Journal of Democracy Khedroob Thondup Geopolitics

Taiwan Needs More Than Democracy to Defend Itself

Taiwan held its first direct presidential election thirty years ago. Today it faces relentless threats from the Chinese Communist Party. Its best defense is self-deterrence. Source

Read at Journal of Democracy

arXiv (research papers) Alexandre Cristovão Maiorano SI

Evaluating Prompt Injection Defenses for Educational LLM Tutors: Security-Usability-Latency Trade-offs

Educational LLM tutors face a core AI alignment challenge: they must follow user intent while preserving pedagogical constraints and safety policies. We present an evaluation methodology for prompt-injection defenses in this setting, showing that guardrail design entails explicit trade-offs among adversarial robustness, benign-task usability, and response latency. We evaluate a domain-specific multi-layer safeguard pipeline combining deterministic pattern filters, structural…

Read at arXiv (research papers)

Stimson Center Communications Geopolitics

North Africa Regional Outlook: March 25, 2026

From Sahel security recalibration to Iran war aftershocks, North Africa faces energy shocks, fiscal strain, and intensifying competition over influence.

Read at Stimson Center

arXiv (research papers) Vincent Siu SI

A Framework for Formalizing LLM Agent Security

Security in LLM agents is inherently contextual. For example, the same action taken by an agent may represent legitimate behavior or a security violation depending on whose instruction led to the action, what objective is being pursued, and whether the action serves that objective. However, existing definitions of security attacks against LLM agents often fail to capture this contextual nature. As a result, defenses face a fundamental utility-security tradeoff: applying…

Read at arXiv (research papers)

Latinoamérica21 Sahasranshu Dash | María Isabel Puerta Riera Geopolitics

The aftermath of Maduro and the dangers of “rapid” transitions: What Latin America can learn from the Middle East and Asia

After Maduro’s fall, Venezuela faces the dilemma of abrupt transitions: how to prevent the end of authoritarianism from giving rise to new forms of unstable or extractive power. O post The aftermath of Maduro and the dangers of “rapid” transitions: What Latin America can learn from the Middle East and Asia apareceu primeiro em Latinoamérica 21.

Read at Latinoamérica21

Journal articles (OpenAlex) M Henriquez, Dorotea López, Megan Ryburn · Journal of Politics in Latin America Geopolitics

Empirical Evidence on Gendered Barriers in Chile's Trade Negotiations

Although debates on gender and trade have advanced significantly in recent years, there is still limited empirical evidence on the participation of women in trade negotiations, particularly with regard to the barriers they face and their capacity to exert substantive influence within negotiation structures. Chile is an important case study for examining gendered dynamics in trade negotiations due to its long-standing and highly institutionalised trade policy. It maintains…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Deborah Manzi, Francesco Calderoni · Journal of Criminal Justice Geopolitics

A simulation approach to assess law enforcement intervention impact: How the efficiency/security trade-off affects drug trafficking networks' resilience to arrests

Despite extensive research on criminal group disruption, little is known about how alternative efficiency/security configurations within drug trafficking organizations (DTOs) shape resilience under sustained law enforcement pressure. DTOs face a core trade-off: efficiency—achieved through coordination and information sharing—increases exposure, while security—based on compartmentalization and trust—constrains flexibility and scale. Although widely theorized, this trade-off…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Elias Isaksson, Alexander Paulsson · Urban Studies Geopolitics

Slowing down: Degrowth and the limits to urban mobility

Cities worldwide face immense challenges in transitioning to a sustainable future. While being structurally and politically bound to continuous growth, the striving for a constant increase in production and consumption puts enormous pressure on our planet and its ecosystems. Degrowth has been proposed as a pathway to solving this dilemma. Although scholarly attention to urban degrowth has expanded, a central aspect of cities remains partly unexplored: mobility. Urban…

Read at Journal articles (OpenAlex)

Latinoamérica21 Inés Amézaga Feijoo Geopolitics

Democracy and Disorder

In a world that is reorganizing itself beyond the bounds of rules and oversight, democracies face the urgent challenge of resisting the advance of authoritarian leadership without renouncing their own limits. O post Democracy and Disorder apareceu primeiro em Latinoamérica 21.

Read at Latinoamérica21

Journal articles (OpenAlex) Thomas Zeitzoff · Terrorism and Political Violence Geopolitics

Where’s the Sabotage? Repression, Recruitment, and Tactical Evolution in the U.S. Radical Environmental Movement

Why did the U.S. radical environmental movement shift away from tactics like arson and sabotage in the mid-2000s? And given the rising threat of climate change, why haven’t these more contentious tactics made a comeback? This puzzle is particularly significant to scholars of terrorism because radical environmental groups like the Earth Liberation Front appear in major terrorism databases, and faced heavy repression from the government, despite causing no fatalities. I argue…

Read at Journal articles (OpenAlex)

Asia Times Thomas P Cavanna South China Sea

US missile deliveries to Philippines are pointless and escalatory

On February 17, the United States announced it would soon deploy additional advanced missiles to the northern Philippines, a part of the archipelago that faces Taiwan. This decision, reflecting the Trump administration’s “iron-clad commitment” to Manila’s security, aims to deter Chinese aggression as Beijing expands its military power and gray-zone warfare in Southeast Asia. However,

Read at Asia Times

Journal articles (OpenAlex) Tapiwa Patson Sisimayi, James Tauya Muperi, Sandra Vuyiswa · Feminist Encounters A Journal of Critical Studies in Culture and Politics Geopolitics

Barriers to women's political participation in Zimbabwe: Analysing structural challenges and pathways to gender equality between 2013 and 2023

The article examines barriers and challenges women face in active political participation within major political parties in Zimbabwe, focusing on the policies and mechanisms that inhibit their involvement while offering recommendations for improvement. Data was collected through interviews and focus groups with 20 purposively and conveniently sampled participants from Harare and Bulawayo, along with critical documentary analysis. The data was analysed methodically using…

Read at Journal articles (OpenAlex)

Dialogue Earth (formerly Diálogo Chino) Rodrigo Rodríguez Geopolitics

Colombia’s green hydrogen ambitions face challenges

Gustavo Petro’s government is making moves to become a major exporter of green hydrogen, but can it be done?

Read at Dialogue Earth (formerly Diálogo Chino)

arXiv (research papers) Gautam Siddharth Kashyap SI

When the Model Said 'No Comment', We Knew Helpfulness Was Dead, Honesty Was Alive, and Safety Was Terrified

Large Language Models (LLMs) need to be in accordance with human values-being helpful, harmless, and honest (HHH)-is important for safe deployment. Existing works use Supervised Fine-Tuning (SFT) and Mixture-of-Experts (MoE) to align LLMs. However, these works face challenges in multi-objective settings, such as SFT leading to interference between conflicting objectives, while MoEs suffer from miscalibrated routing. We term this failure mode Axis Collapse, marked by (1)…

Read at arXiv (research papers)

arXiv (research papers) Mingrui Liu SI

RedVisor: Reasoning-Aware Prompt Injection Defense via Zero-Copy KV Cache Reuse

Large Language Models (LLMs) are increasingly vulnerable to Prompt Injection (PI) attacks, where adversarial instructions hidden within retrieved contexts hijack the model's execution flow. Current defenses typically face a critical trade-off: prevention-based fine-tuning often degrades general utility via the "alignment tax", while detection-based filtering incurs prohibitive latency and memory costs. To bridge this gap, we propose RedVisor, a unified framework that…

Read at arXiv (research papers)

Journal articles (OpenAlex) Soon-ok Shin, Chris Alden, Richard A. Higgott · Global Policy Geopolitics

Re‐Imagining Middle Powers in an Age of Global Transformation

ABSTRACT The post–1945 international order is in flux. Trump's return to the White House in 2025 marks a critical juncture, reinforcing a structural transformation characterised by intensifying US–China rivalry, the weaponisation of economic statecraft and the fragmentation of multilateralism. Middle powers—positioned at the intersection of Washington's security imperatives and Beijing's economic gravitational pull—face mounting strategic challenges, but with the requisite…

Read at Journal articles (OpenAlex)

arXiv (research papers) Junda Lin SI

VIGIL: Defending LLM Agents Against Tool Stream Injection via Verify-Before-Commit

LLM agents operating in open environments face escalating risks from indirect prompt injection, particularly within the tool stream where manipulated metadata and runtime feedback hijack execution flow. Existing defenses encounter a critical dilemma as advanced models prioritize injected rules due to strict alignment while static protection mechanisms sever the feedback loop required for adaptive reasoning. To reconcile this conflict, we propose \textbf{VIGIL}, a framework…

Read at arXiv (research papers)

arXiv (research papers) Qiang Yu SI

Defense Against Indirect Prompt Injection via Tool Result Parsing

As LLM agents transition from digital assistants to physical controllers in autonomous systems and robotics, they face an escalating threat from indirect prompt injection. By embedding adversarial instructions into the results of tool calls, attackers can hijack the agent's decision-making process to execute unauthorized actions. This vulnerability poses a significant risk as agents gain more direct control over physical environments. Existing defense mechanisms against…

Read at arXiv (research papers)

arXiv (research papers) Zhiyuan Chang SI

Know Thy Enemy: Securing LLMs Against Prompt Injection via Diverse Data Synthesis and Instruction-Level Chain-of-Thought Learning

Large language model (LLM)-integrated applications have become increasingly prevalent, yet face critical security vulnerabilities from prompt injection (PI) attacks. Defending against PI attacks faces two major issues: malicious instructions can be injected through diverse vectors, and injected instructions often lack clear semantic boundaries from the surrounding context, making them difficult to identify. To address these issues, we propose InstruCoT, a model enhancement…

Read at arXiv (research papers)

Older stories