Results for “App”

Simon Willison SI

The Normalization of Deviance in AI

The Normalization of Deviance in AI This thought-provoking essay from Johann Rehberger directly addresses something that I’ve been worrying about for quite a while: in the absence of any headline-grabbing examples of prompt injection vulnerabilities causing real economic harm, is anyone going to care? Johann describes the concept of the “Normalization of Deviance” as directly applying to this question. Coined by Diane Vaughan, the key idea here is that organizations that get…

Read at Simon Willison

OWASP GenAI Security Project John Sotiropoulos Security

OWASP Top 10 for Agentic Applications – The Benchmark for Agentic Security in the Age of Autonomous AI

Introducing the OWASP Top 10 for Agentic AI Applications - our community’s actionable framework for securing autonomous, tool-using AI systems. Built at global scale informed by real incidents, and based on our work already adopted across industry, this release marks a pivotal moment in turning insight into action, and advancing the security of Agentic Applications at the pace of innovation.

Read at OWASP GenAI Security Project

arXiv (research papers) Reachal Wang SI

ObliInjection: Order-Oblivious Prompt Injection Attack to LLM Agents with Multi-source Data

Prompt injection attacks aim to contaminate the input data of an LLM to mislead it into completing an attacker-chosen task instead of the intended task. In many applications and agents, the input data originates from multiple sources, with each source contributing a segment of the overall input. In these multi-source scenarios, an attacker may control only a subset of the sources and contaminate the corresponding segments, but typically does not know the order in which the…

Read at arXiv (research papers)

arXiv (research papers) Yinan Zhong SI

Attention is All You Need to Defend Against Indirect Prompt Injection Attacks in LLMs

Large Language Models (LLMs) have been integrated into many applications (e.g., web agents) to perform more sophisticated tasks. However, LLM-empowered applications are vulnerable to Indirect Prompt Injection (IPI) attacks, where instructions are injected via untrustworthy external data sources. This paper presents Rennervate, a defense framework to detect and prevent IPI attacks. Rennervate leverages attention features to detect the covert injection at a fine-grained token…

Read at arXiv (research papers)

Journal articles (OpenAlex) María Vásquez Callo-Müller, Franziska Sucker · International Data Privacy Law Geopolitics

Evolving approaches to cross-border data flows: Latin American and African perspectives

This article examines the governance of cross-border data flows (CBDF) within Latin American and African regional economic integration frameworks, focusing on the Pacific Alliance (PA), Mercosur, and the African Continental Free Trade Area (AfCFTA). Through a comparative legal analysis of the PA’s First Amending Protocol, Mercosur’s E-Commerce Agreement, and the recent AfCFTA’s Protocol on Digital Trade (DTP), we assess how regional frameworks regulate CBDF amid evolving…

Read at Journal articles (OpenAlex)

arXiv (research papers) M Zeeshan SI

Chameleon: Adaptive Adversarial Agents for Scaling-Based Visual Prompt Injection in Multimodal AI Systems

Multimodal Artificial Intelligence (AI) systems, particularly Vision-Language Models (VLMs), have become integral to critical applications ranging from autonomous decision-making to automated document processing. As these systems scale, they rely heavily on preprocessing pipelines to handle diverse inputs efficiently. However, this dependency on standard preprocessing operations, specifically image downscaling, creates a significant yet often overlooked security…

Read at arXiv (research papers)

arXiv (research papers) Eranga Bandara SI

ASTRIDE: A Security Threat Modeling Platform for Agentic-AI Applications

AI agent-based systems are becoming increasingly integral to modern software architectures, enabling autonomous decision-making, dynamic task execution, and multimodal interactions through large language models (LLMs). However, these systems introduce novel and evolving security challenges, including prompt injection attacks, context poisoning, model manipulation, and opaque agent-to-agent communication, that are not effectively captured by traditional threat modeling…

Read at arXiv (research papers)

arXiv (research papers) Chenlin Xu SI

Boundary-Aware Test-Time Adaptation for Zero-Shot Medical Image Segmentation

Due to the scarcity of annotated data and the substantial computational costs of model, conventional tuning methods in medical image segmentation face critical challenges. Current approaches to adapting pretrained models, including full-parameter and parameter-efficient fine-tuning, still rely heavily on task-specific training on downstream tasks. Therefore, zero-shot segmentation has gained increasing attention, especially with foundation models such as SAM demonstrating…

Read at arXiv (research papers)

arXiv (research papers) Omar Farooq Khan Suri SI

Securing Large Language Models (LLMs) from Prompt Injection Attacks

Large Language Models (LLMs) are increasingly being deployed in real-world applications, but their flexibility exposes them to prompt injection attacks. These attacks leverage the model's instruction-following ability to make it perform malicious tasks. Recent work has proposed JATMO, a task-specific fine-tuning approach that trains non-instruction-tuned base models to perform a single function, thereby reducing susceptibility to adversarial instructions. In this study, we…

Read at arXiv (research papers)

Journal articles (OpenAlex) Bořivoj Nachtmann · CyberOrient Geopolitics

Finishing the Fight for Freedom: The Islamic Emirate of Afghanistan's Strategic Communication of the “Freedom Jihad” Narrative Frame on X

Abstract This article examines the strategic communication of the Islamic Emirate of Afghanistan (IEA) on the social media platform X (formerly Twitter) between November 2022 and August 2023. Specifically, it focuses on the official spokesperson, Zabihullah Mujahid, whose account is the central node of the IEA's X network. Drawing on Nicole Ball's framework of post‐conflict state reconstruction and utilizing the approach of Teun A. van Dijk's critical‐political discourse…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Samuel Lee Yu Sum · Theoria South China Sea

‘Un-Islandic’ Island-Making

Abstract This article examines China's artificial island-building in the South China Sea as a form of ‘un-islandic’ violence that undermines ecological and relational principles central to archipelagic thinking. Drawing on Carl Schmitt's theory of nomos and land appropriation, it analyses how Chinese political theorists and state planners have embraced Schmittian concepts to justify territorial expansion and sovereign authority. It argues that the transformation of coral…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Samuel Lee Yu Sum, Miguel Vatter · Theoria South China Sea

Archipelagic Thinking and Its Political Implications in the Western Pacific

This special issue aims to illustrate how the emerging approach of archipelagic thinking can contribute to political theory and international relations theory, and it attempts to grapple with the geopolitical and ontological complexities of the South China Sea and Western Pacific. Archipelagic thinking is an emerging interdisciplinary approach in island studies, political anthropology and political ecology. The interest of archipelagic thinking for political theorising rests…

Read at Journal articles (OpenAlex)

arXiv (research papers) K. J. Kevin Feng SI

On the Regulatory Potential of User Interfaces for AI Agent Governance

AI agents that take actions in their environment autonomously over extended time horizons require robust governance interventions to curb their potentially consequential risks. Prior proposals for governing AI agents primarily target system-level safeguards (e.g., prompt injection monitors) or agent infrastructure (e.g., agent IDs). In this work, we explore a complementary approach: regulating user interfaces of AI agents as a way of enforcing transparency and behavioral…

Read at arXiv (research papers)

Journal articles (OpenAlex) Medeu Kurmangali, Talgat Akimzhanov, Kanat Kazhibaev et al. · Journal of Cultural Analysis and Social Change Geopolitics

Kazakhstan's Drug Policy and Practice Today: Is the Reactive Approach of Fighting the Consequences of Drug Trafficking while Ignoring its Causes Changing?

Kazakhstan is facing an increase in the smuggling of precursors from Russia and an increase in domestic production of synthetic drugs. Young people are increasingly involved in drug trafficking, both as consumers and distributors. The spread of Internet technologies has facilitated the online sale of drugs, and the patterns of covert trade that originated in Russia are now spreading to Kazakhstan. Qualitative research methods are used, including analysis of legislative…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Maria-Beatrice Berna · Journal of Artificial Intelligence & Cloud Computing SI

Human Rights as the Underpinning Legal Constraint in the Use and Deployment of Lethal Autonomous Weapons (LAWS): LAWS Consequences on Core Human Rights, The Urgent Need for Legal Accountability Mechanisms, and the Rationale for a Comprehensive International Regulatory Framework

The conceptual foundation of artificial intelligence can be traced to the Three Laws of Robotics, formulated by Isaac Asimov in his 1950 science fiction literature. These principles articulated a vision of intelligent systems designed to serve human advancement and societal progress. All the Three Laws of Robotics build upon upholding the unique essence of human existence; following the human rights based approach on artificial intelligence, it is safe to presume that it was…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Junhua Zhu, Elina Sinkkonen, Mikael Mattlin · Telecommunications Policy Geopolitics

Strategic technology competition revisited: A National Innovation System rationale for China's artificial intelligence standardisation strategy

This article approaches the question of China’s ability to set international standards in the field of artificial intelligence (AI) by analysing the co-evolution of China’s national innovation system (NIS) and standardisation strategy. In theory, emerging technologies, such as AI, where global norms and standard institutions’ governance are yet to be established, offer greater possibilities for latecomers to set international standards. As areas of AI are general-purpose…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Elizabeth A. Tomsich, Garen J. Wintemute · Injury Epidemiology Geopolitics

Militancy in the military: military service and support for political violence and right-wing extremism

BACKGROUND: Political violence constitutes an increasing threat to individual and population-level health in the United States, with military service identified as a potential risk factor. The current study examines the association between military service, combat experience, and support for and willingness to engage in political violence and approval of extremist organizations and movements. METHODS: A nationally representative sample of 12,947 US adult members of the Ipsos…

Read at Journal articles (OpenAlex)

arXiv (research papers) Kaiyuan Zhang SI

BrowseSafe: Understanding and Preventing Prompt Injection Within AI Browser Agents

The integration of artificial intelligence (AI) agents into web browsers introduces security challenges that go beyond traditional web application threat models. Prior work has identified prompt injection as a new attack vector for web agents, yet the resulting impact within real-world environments remains insufficiently understood. In this work, we examine the landscape of prompt injection attacks and synthesize a benchmark of attacks embedded in realistic HTML payloads…

Read at arXiv (research papers)

CSIS Asia Maritime Transparency Initiative Bich Tran South China Sea

The Consequences of Shelving the South China Sea Disputes

For years, China has promoted the idea of “shelving disputes” in the South China Sea and focusing on areas of cooperation. Southeast Asian claimants—Brunei, Malaysia, the Philippines, and Vietnam—have been receptive to this approach to varying degrees. While this strategy brings significant economic benefits, it carries substantial long-term risks for their territorial claims and regional

Read at CSIS Asia Maritime Transparency Initiative

Journal articles (OpenAlex) Tran Minh Duc · Journal of Cultural Analysis and Social Change Geopolitics

Constructing Boundaries through Terminology: How ‘Culture’ and ‘Creativity’ Shape Inter-Ministerial Governance in Vietnam’s Cultural and Creative Industries Policy

The focus of this piece is to investigate how “culture” and “creativity” inform and contour the policies relating to Cultural and Creative Industries (CCIs) in Vietnam. The author takes a qualitative approach to the study by analyzing meeting minutes and official documents, conducting interviews, and reviewing relevant media sources. The findings of this study indicate that the language of policy functions as a boundary tool, and as such, shapes the jurisdictional…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Nguyen Hung Vuong, Vo Ha Chi · Journal of Cultural Analysis and Social Change South China Sea

Undersea Cables in the South China Sea: A New Frontier of Us-China Strategic Competition and The Dilemma for Southeast Asian Countries

This research paper examines the strategic competition between the United States and China over undersea fiber-optic infrastructure in the South China Sea, a region rapidly emerging as a new frontier in the global digital order. Drawing on the theory of “Weaponized Interdependence”, the article demonstrates how these undersea cables have become politicized, serving as instruments for exerting systemic influence, enacting strategic denial, and applying soft coercion. While…

Read at Journal articles (OpenAlex)

arXiv (research papers) Steven Peh SI

Prompt Fencing: A Cryptographic Approach to Establishing Security Boundaries in Large Language Model Prompts

Large Language Models (LLMs) remain vulnerable to prompt injection attacks, representing the most significant security threat in production deployments. We present Prompt Fencing, a novel architectural approach that applies cryptographic authentication and data architecture principles to establish explicit security boundaries within LLM prompts. Our approach decorates prompt segments with cryptographically signed metadata including trust ratings and content types, enabling…

Read at arXiv (research papers)

arXiv (research papers) Qingsong He SI

Z-Space: A Multi-Agent Tool Orchestration Framework for Enterprise-Grade LLM Automation

Large Language Models can break through knowledge and timeliness limitations by invoking external tools within the Model Context Protocol framework to achieve automated execution of complex tasks. However, with the rapid growth of enterprise-scale MCP services, efficiently and accurately matching target functionalities among thousands of heterogeneous tools has become a core challenge restricting system practicality. Existing approaches generally rely on full-prompt…

Read at arXiv (research papers)

Journal articles (OpenAlex) Efe Peker, Emily Laxer, Rémi Vivès · Canadian Review of Sociology/Revue canadienne de sociologie Geopolitics

A Right‐Wing Populist Turn in the Conservative Party of Canada? Continuities and Ruptures Under the Leadership of Pierre Poilievre (2022–2025)

Since his election as leader of the Conservative Party of Canada (CPC) in 2022, Pierre Poilievre has been associated with populism in media and political discourse, with implicit and explicit comparisons to Donald Trump. This article investigates the validity of such assessments by applying "complex" theories of populism, which conceptualize populism as an interplay of ideology, communication style, and organizational strategy. Drawing on a qualitative content analysis from…

Read at Journal articles (OpenAlex)

arXiv (research papers) Aram Vardanyan SI

Building Browser Agents: Architecture, Security, and Practical Solutions

Browser agents enable autonomous web interaction but face critical reliability and security challenges in production. This paper presents findings from building and operating a production browser agent. The analysis examines where current approaches fail and what prevents safe autonomous operation. The fundamental insight: model capability does not limit agent performance; architectural decisions determine success or failure. Security analysis of real-world incidents reveals…

Read at arXiv (research papers)

CSIS Asia Maritime Transparency Initiative Vu Hai Dang South China Sea

Scarborough Peace Park: The Road to Peace and Sustainability in the South China Sea

Introduction On September 10, China approved the establishment of a nature reserve at Scarborough Shoal, also known as Bajo de Masinloc in Tagalog or Huangyan Dao in Chinese, a coral atoll whose sovereignty is under dispute between China and the Philippines. According to Beijing, the reserve covers approximately one acre on the northeastern side of

Read at CSIS Asia Maritime Transparency Initiative

Asia Times Imran Khalid South China Sea

Time for a US Missile Mutual Assurance Pact for Asia

In the shadow of Russia’s grinding war in Ukraine, Europe is stirring from a long slumber on defense. On November 14, the United States approved a US$3.5 billion arms deal that will arm Germany with 173 Standard Missile-6 interceptors and 577 Standard Missile-2s, the largest such transaction in years. These are not just weapons. They

Read at Asia Times