Results for “Again”

Journal articles (OpenAlex) Damilola Waliyullahi. Abimbola, Akin Ademuyiwa, Tokunbo Salako · The International Spectator Geopolitics

Temporal Power and Coercive Diplomacy in Russia’s War against Ukraine

Temporal manipulation has functioned as a strategic instrument of coercive diplomacy in Russia's conduct of the war in Ukraine, rather than merely as a by-product of military stalemate. Through the deliberate management of negotiation rhythms, deadlines and diplomatic procedures, Russia has shaped bargain dynamics throughout the conflict. Drawing on a qualitative analysis of negotiation timelines, diplomatic procedures and public diplomatic performances, three interrelated…

Read at Journal articles (OpenAlex)

InSight Crime Deborah Bonello Geopolitics

On the Radar: Mexico’s Criminal Pressure Points: Drugs, Corruption, and CJNG Fuel Theft

On the Radar: Mexico’s Criminal Pressure Points: Drugs, Corruption, and CJNG Fuel Theft Who is winning the drug war - governments or drugs? Is Mexico's ruling party cracking under pressure from President Trump and organized crime? What is in the crosshairs of the latest OFAC sanctions against the Jalisco Cartel target? We discuss in this week's On the Radar.

Read at InSight Crime

Article 19 natashas

Malaysia: Cybercrimes Bill 2026 is a threat to free expression and privacy

ARTICLE 19, the Centre for Independent Journalism (CIJ), and Sinar Project are concerned about the threat posed by the Malaysian Cybercrimes Bill 2026 to the right to freedom of expression online, as well as to media, journalists and broadcast organisations in Malaysia. Although Malaysia is a signatory to the United Nations Convention against Cybercrime, the

Read at Article 19

arXiv (research papers) Mitchell Hermon SI

Security--Fidelity Tradeoffs: The Hidden Cost of Prompt Injection Defense

We identify a security-fidelity tradeoff in defending LLMs against indirect prompt injection: defenses resist injected instructions largely by suppressing untrusted text, which corrupts tasks that must preserve it, such as translation and document editing. Attack-success metrics cannot see this, because a model that ignores an injection and one that faithfully processes it as data score identically. We introduce SecFid, a benchmark built so that executing an injection…

Read at arXiv (research papers)

Article 19 natashas

Iran: Human rights must be at the centre of engagement with Islamic Republic

As hostilities give way to diplomacy and governments across the world begin repositioning their relationships with the Islamic Republic of Iran, ARTICLE 19, as part of the Impact Iran Coalition, expresses its profound concern that the human rights crisis facing the people of Iran is once again being sidelined. This pattern is neither new nor

Read at Article 19

The Guardian Syria Maanvi Singh Geopolitics

Trump finds partner in supreme court in his war against immigration

Recent rulings have bolstered president’s pitiless campaign against people of colour fleeing violence and disaster In the US supreme court, it seems Donald Trump has found loyalists in his crusade against immigration and immigrants. In a pair of rulings on Thursday, the court allowed the Trump administration to end humanitarian protections that have granted people from Haiti and Syria the right to live and work legally in the US for more than a decade, and cleared way for…

Read at The Guardian Syria

Phenomenal World Fernando Rugitsky Geopolitics

Drowning in Debt

The upcoming presidential election in Brazil, set to take place in October of this year, will for a third time pit Luís Inácio Lula da Silva’s Workers’ Party against the far-right movement led by Jair Bolsonaro. Lula is running for reelection at the end of his third term as president, campaigning on the back of

Read at Phenomenal World

arXiv (research papers) Praneeth Narisetty SI

Adaptive Evaluation of Out-of-Band Defenses Against Prompt Injection in LLM Agents

Recent work (2024 to 2026) has converged on a strategy for defending tool-using LLM agents against indirect prompt injection: rather than training the model to refuse malicious instructions, enforce security outside the model with a deterministic policy that mediates the agent's actions. Systems such as CaMeL, FIDES, Progent, RTBAS, and FORGE realize this with capabilities, information-flow labels, and reference monitors, and several report near-elimination of attacks on the…

Read at arXiv (research papers)

arXiv (research papers) Poojitha Thota SI

Verifying Intent and Harm: A Unified Defense Against LLM-Generated Threats

Large language models (LLMs) are increasingly deployed in interactive applications, yet they remain vulnerable to adversarial interactions that induce harmful, deceptive, or policy-violating outputs. Existing defenses typically analyze either user prompts or generated outputs, but not both. However, many real-world attacks exploit a separation between adversarial intent expressed in the prompt and actionable harm manifested only in the response. As a result, prompt-only and…

Read at arXiv (research papers)

InSight Crime Jeremy McDermott Geopolitics

The InSight Take: Colombia Turns Right With Abelardo de la Espriella. Here Are the Washington and Security Challenges.

The InSight Take: Colombia Turns Right With Abelardo de la Espriella. Here Are the Washington and Security Challenges. The winner of Colombia’s presidential elections, Abelardo de la Espriella, promised to stop negotiating with guerrilla armies and to unleash the country’s experienced security forces against criminal groups there during his political campaign. But will the Trump-aligned president-elect keep his promises when he takes office in August? And will his plan help…

Read at InSight Crime

arXiv (research papers) Yang Gao Veyon Solutions SI

How Reliable Is Your Jailbreak Judge? Calibration and Adversarial Robustness of Automated ASR Scoring

Almost every paper on LLM jailbreaks and prompt injection reports an attack-success rate (ASR), and that number is assigned not by people but by an automated judge: either a safety classifier trained for the task, or a general chat model prompted to grade. The judge is rarely checked. We check it. Using 596 human-labeled completions from the HarmBench classifier validation set, we compare the two judge families against human majority votes and then attack them. The two…

Read at arXiv (research papers)

arXiv (research papers) Yedidel Louck SI

Securing LLM-Agent Long-Term Memory Against Poisoning: Non-Malleable, Origin-Bound Authority with Machine-Checked Guarantees

LLM agents increasingly rely on persistent long-term memory, which creates a critical vulnerability that we study here: memory poisoning. An adversary can store untrusted content in one session that later steers a consequential action, such as a payment, a setting change, or data exfiltration, in a future session. Existing defenses base a memory item's authority to act on either its content (detection or trust-scoring) or its derivation history (lineage). We show that both…

Read at arXiv (research papers)

arXiv (research papers) Xingwei Zhong SI

DE-FIVE: Detecting Malicious Image Prompts via Fourier Features and Image Vector Embeddings

Vision language models (VLMs) employ both visual and textual modalities to enable advanced vision-language inference. However, incorporating visual modalities expands the attack surface of VLMs, making them more susceptible to security threats such as adversarial perturbations and indirect prompt injection, wherein crafted malicious image prompts can elicit unintended model outputs. Existing defense methods against malicious image prompts remain insufficient as they…

Read at arXiv (research papers)

Journal articles (OpenAlex) Jenny Ufer, Felix Windegger, Matthias Schmelzer et al. · New Global Studies Geopolitics

Introduction. Global (Inter)dependencies and Degrowth: Towards an Analytical Framework

Abstract So far, degrowth scholarship remains largely shaped by a form of methodological localism or nationalism that risks reproducing current asymmetric and colonial relations on the global scale. Against this background, our introduction to the forum argues that degrowth can only adequately be theorized through the lens of global (inter)dependencies: structurally reproduced relations of dependence and mutual entanglement that operate across national borders and generate…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Pınar Bedirhanoğlu · Capital & Class Geopolitics

Towards an Open and Political Marxist theory of the state: Levelling Simon Clarke’s ‘rule of money’ and Ellen Meiksins Wood’s ‘market imperative’

The state became once again a central question in Marxist studies. This interest displays a specific engagement with the state as explanandum – as a phenomenon to be explained in itself – in contrast to an interest in the state simply as explanans , or in terms of its role and functions within neoliberal capitalism. This might be due to the urgent concern to understand whether there are any class limits to the extraordinary and radical political transformations taking place…

Read at Journal articles (OpenAlex)

arXiv (research papers) Shanghao Shi SI

Think Twice Before You Act: Protecting LLM Agents Against Tool Description Poisoning via Isolated Planning

The integration of external tools has substantially expanded the capabilities of large language model (LLM) agents, but it also introduces new attack surfaces beyond prompt injection. In particular, cross-tool description poisoning can manipulate planner-visible tool metadata to steer an agent's trajectory, even if the poisoned tool itself is never chosen. To understand the effectiveness of existing defenses against this emerging threat, we first evaluate several…

Read at arXiv (research papers)

arXiv (research papers) Reza Soosahabi SI

Analyzing Defensive Misdirection Against Model-Guided Automated Attacks on Agentic AI Systems

Agentic AI systems increasingly rely on language-model components to interpret instructions, process external data, invoke tools, and coordinate with other agents. These capabilities make prompt-injection and jailbreak attacks more consequential, especially as attackers adopt model-guided automation to scale probing, prompt refinement, and response evaluation. This work analyzes the resulting attack-defense setting through a probabilistic model of a target system, its…

Read at arXiv (research papers)

Journal articles (OpenAlex) Xavier Lafrance · Capital & Class Geopolitics

Social ontology in open Marxism and political Marxism: From alienation, to exploitation and back again

This article compares how Open Marxism and Political Marxism, respectively, mobilize the concepts of social form and of social property relations to produce critiques of the fetishized social ontology of mainstream or orthodox Marxism(s). Focusing on the works of Simon Clarke, Robert Brenner and Ellen Meiksins Wood, I show how these thinkers develop strikingly convergent readings of Marx’s critique of political economy as a rejection of an ontologically separate and asocial…

Read at Journal articles (OpenAlex)

arXiv (research papers) Gulshan Saleem SI

A Layered Security Framework Against Prompt Injection in RAG-Based Chatbots

Prompt injection is ranked as the most critical vulnerability in large language model (LLM) deployments by the OWASP Top 10 for LLM Applications, yet existing defenses operate at isolated pipeline stages and remain incomplete. Input filters cannot inspect retrieved documents, while output monitors cannot prevent malicious payloads from reaching the model. Consequently, retrieval-augmented generation (RAG) chatbots remain vulnerable to indirect injection, where a poisoned…

Read at arXiv (research papers)

arXiv (research papers) J Alex Corll SI

Amplify, Don't Create: Temporal Accumulation for Slow-Burn Prompt Injection

Most prompt-injection detectors score a single event or message. Control-plane attacks against tool-using agents can instead distribute weak directives across a trajectory while keeping each event below threshold. We test whether a proxy-side temporal accumulator recovers this slow-burn signal by reducing frozen per-event scores to peak and CUSUM persistence statistics. To avoid circularity, grafts are generated against a held-out autoregressive cloaking target and then…

Read at arXiv (research papers)

arXiv (research papers) Po-Han Cheng SI

CodeSentinel: A Three-Layer Defense Against Indirect Prompt Injection in Code Contexts

Code large language models increasingly retrieve external code context from repositories, documentation, issue threads, and coding-agent environments, creating an indirect prompt-injection surface where attackers hide instructions in comments, strings, identifiers, or decoy code. We propose CodeSentinel, a three-layer inference-time sanitizer. It uses Tree-sitter to extract high-risk model-facing CST nodes, then combines syntax-guided pre-filtering, CST-guided Dynamic…

Read at arXiv (research papers)

arXiv (research papers) Laxmipriya Ganesh Iyer SI

The Gate Is Only as Honest as Its Contracts: ContractGuard for the Contract Layer of Risk-Aware Causal Gating

Risk-Aware Causal Gating (RACG) defends tool-augmented LLM agents against indirect prompt injection by removing dangerous tools from the agent's visible action space, so that even a fully injection-compliant agent cannot call a tool it cannot see. We make three points. First, this structural guarantee does not eliminate the trust assumption behind safe tool use; it relocates it into the integrity of the tool contracts -- declared preconditions, effects, risk, and…

Read at arXiv (research papers)

The Guardian Somalia Mark Townsend and Mohamed Gabobe in Mogadishu

Killed walking home from school: why did Somali children become targets of US drone strikes?

Six months ago, at least 12 people, including eight children, died during a US attack. The US has never admitted the civilian deaths. Here, the Guardian pieces together what happened that day Explainer: Why is the US bombing Somalia – and who are the airstrikes killing? They had just settled down for breakfast when the noise came. Some paused in the eating of their slow-cooked beans – cambuulo – spooked by the haunting high-pitched hum. Others pressed their faces against…

Read at The Guardian Somalia

arXiv (research papers) David Huang SI

Rapid Poison: Practical Poisoning Attacks Against the Rapid Response Framework

The Rapid Response (RR) framework, deployed in production systems, including Anthropic's ASL-3 safeguards, continuously improves jailbreak-detection classifiers. When new jailbreaks emerge that bypass these classifiers, Rapid Response generates synthetic variants for training, helping the model generalize from the new attacks and quickly adapt. We reveal that prompt injection can infiltrate this pipeline to deliver poisoned samples into the classifier's training set…

Read at arXiv (research papers)

Inquirer Global Nation dsanchez South China Sea

NMC defends Teodoro vs China sanctions, upholds rights in West PH Sea

MANILA, Philippines — The Marcos-created National Maritime Council (NMC) has pushed back against the sanctions imposed by China on Defense Secretary Gilberto ”Gibo” Teodoro Jr. and his family. In a statement dated June 13 but released only on Sunday, the NMC expressed its support for Teodoro, saying he “has consistently carried out his duties in accordance

Read at Inquirer Global Nation

arXiv (research papers) Lipeng He SI

Defending against Adaptive Prompt Injection Attacks via Reasoning-enabled Task Alignment

Indirect prompt injection attacks hijack LLM-based agents by embedding malicious instructions in third-party data that the agent retrieves during task execution. Existing defenses report near-zero attack success rate on static benchmarks, yet recent adaptive evaluations show that these results collapse once the attacker is allowed to optimize against the deployed defense. In this work, we trace this collapse to two failure modes. First, existing defense methods are confined…

Read at arXiv (research papers)

arXiv (research papers) Xinhang Ma SI

AutoDojo: A Generative Benchmark for Evaluating Prompt Injection Defenses in LLM Agents

Indirect prompt injection (IPI) is a major security threat to LLM-powered agents. Thus, a growing body of work has proposed a variety of defensive approaches against IPI which can be grouped into three broad categories: prompt-level, filter-based, and system-level designs. However, commonly used benchmarks for evaluating defense, such as AgentDojo, are \emph{inherently static}, generating a fixed distribution of IPI attacks. Consequently, a defense can score well on them…

Read at arXiv (research papers)

arXiv (research papers) Yuguang Zhou SI

From Shield to Target: Denial-of-Service Attacks on LLM-Based Agent Guardrails

LLM-based guardrails have emerged as a highly effective defense against prompt injection and jailbreak attacks in autonomous agents. However, we reveal that the very reasoning and task-following capabilities enabling this protection introduce a novel vulnerability: attackers can inject crafted data to trap the guardrail in extended reasoning loops, effectuating a systematic denial-of-service (DoS) attack. To systematically expose this threat, we design a beam-search…

Read at arXiv (research papers)

The Diplomat Sebastian Strangio South China Sea

Philippine Defense Secretary Vows to Defend Nation’s Interests After China Sanctions

The country's Department of Foreign Affairs said that the sanctions against Gilberto Teodoro Jr. are “an unfriendly act" that "further complicates" bilateral relations.

Read at The Diplomat

arXiv (research papers) Timothy McAllister SI

Smarter Saboteurs, Better Fixers: Scaling & Security in Linear Multi-Agent Workflows

As LLM-based multi-agent systems (MAS) are deployed in the wild, the resilience of their collaboration structures against adversarial compromise becomes a critical safety concern. Attackers may leverage prompt-injection or jailbreaking to sabotage individual agents within MAS workflows, but the interaction between model scaling and system-level resilience remains poorly understood. This paper investigates how model scale affects the security of linear multi-agent workflows…

Read at arXiv (research papers)

Older stories