Results for “Meta”

arXiv (research papers) Al Muhit Muhtadi SI

MIPIAD: Multilingual Indirect Prompt Injection Attack Defense with Qwen -- TF-IDF Hybrid and Meta-Ensemble Learning

Indirect prompt injection remains a persistent weakness in retrieval-augmented and tool-using LLM systems, and the problem becomes harder to characterise in multilingual settings. We present MIPIAD, a defense framework evaluated on English and Bangla that combines a sequence classifier fine-tuned from Qwen2.5-1.5B via LoRA (XLPID), TF-IDF lexical features, and validation-tuned ensembling through late fusion, stacking, and gradient boosting. The framework is evaluated on a…

Read at arXiv (research papers)

War on the Rocks Rashed M. Aba-namay Geopolitics

Why Iran Metabolizes the Pressure that Broke Venezuela

The reality that regime change is not going to happen as a result of this war seems to have settled in at the White House. When American policymakers reflect and wonder why Iran did not react like Venezuela under pressure, they will not just be misreading Iran — they will be misreading how coercive pressure works. Iran’s resilience rests on internal and external pillars that a Venezuela comparison completely misses. Internally, the Supreme Leader’s authority is anchored in a…

Read at War on the Rocks

arXiv (research papers) Dongcheng Zhang SI

Owner-Harm: A Missing Threat Model for AI Agent Safety

Existing AI agent safety benchmarks focus on generic criminal harm (cybercrime, harassment, weapon synthesis), leaving a systematic blind spot for a distinct and commercially consequential threat category: agents harming their own deployers. Real-world incidents illustrate the gap: Slack AI credential exfiltration (Aug 2024), Microsoft 365 Copilot calendar-injection leaks (Jan 2024), and a Meta agent unauthorized forum post exposing operational data (Mar 2026). We propose…

Read at arXiv (research papers)

arXiv (research papers) Qianlong Lan SI

Silent Egress: When Implicit Prompt Injection Makes LLM Agents Leak Without a Trace

Agentic large language model systems increasingly automate tasks by retrieving URLs and calling external tools. We show that this workflow gives rise to implicit prompt injection: adversarial instructions embedded in automatically generated URL previews, including titles, metadata, and snippets, can introduce a system-level risk that we refer to as silent egress. Using a fully local and reproducible testbed, we demonstrate that a malicious web page can induce an agent to…

Read at arXiv (research papers)

Journal articles (OpenAlex) Axel Bastián Poque González, Yunesky Masip Macía · Sustainability Geopolitics

Energy Justice, Critical Minerals, and the Geopolitical Metabolism of the Global Energy Transition: Insights from Copper Extraction in Chile and Peru

The global energy transition (ET) is widely portrayed as a technological shift toward low-carbon systems; however, it also entails profound geopolitical and socio-environmental transformations. While energy justice (EJ) has become a key framework for assessing fairness in energy systems, it seldom incorporates the geopolitical restructuring of material, energy, and economic flows that underpin contemporary transitions. This article develops a geopolitically informed approach…

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)

Journal articles (OpenAlex) Ondřej Rosendorf, Michal Smetana, Marek Albert Vranka et al. · International Studies Quarterly SI

Mind over Metal: Public Opinion on Autonomous Weapons in the United States, Brazil, Germany, and China

Abstract Autonomous weapons exemplify the tension between technology and ethics in modern warfare. While observational studies suggest that the public holds negative attitudes toward the use of these weapons, experimental evidence from the United States shows that their acceptance is context dependent. This suggests that Americans may hold only a weak preference for “mind over metal,” and the aversion observed in polls may not reflect deeply held moral convictions. Yet…

Read at Journal articles (OpenAlex)

arXiv (research papers) Toqeer Ali Syed SI

Toward Trustworthy Agentic AI: A Multimodal Framework for Preventing Prompt Injection Attacks

Powerful autonomous systems, which reason, plan, and converse using and between numerous tools and agents, are made possible by Large Language Models (LLMs), Vision-Language Models (VLMs), and new agentic AI systems, like LangChain and GraphChain. Nevertheless, this agentic environment increases the probability of the occurrence of multimodal prompt injection (PI) attacks, in which concealed or malicious instructions carried in text, pictures, metadata, or agent-to-agent…

Read at arXiv (research papers)

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) 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)