Results for “Again”

arXiv (research papers) Jiani Liu SI

PINA: Prompt Injection Attack against Navigation Agents

Navigation agents powered by large language models (LLMs) convert natural language instructions into executable plans and actions. Compared to text-based applications, their security is far more critical: a successful prompt injection attack does not just alter outputs but can directly misguide physical navigation, leading to unsafe routes, mission failure, or real-world harm. Despite this high-stakes setting, the vulnerability of navigation agents to prompt injection…

Read at arXiv (research papers)

arXiv (research papers) Anirudh Sekar SI

Zero-Shot Embedding Drift Detection: A Lightweight Defense Against Prompt Injections in LLMs

Prompt injection attacks have become an increasing vulnerability for LLM applications, where adversarial prompts exploit indirect input channels such as emails or user-generated content to circumvent alignment safeguards and induce harmful or unintended outputs. Despite advances in alignment, even state-of-the-art LLMs remain broadly vulnerable to adversarial prompts, underscoring the urgent need for robust, productive, and generalizable detection mechanisms beyond…

Read at arXiv (research papers)

OnCuba News OnCuba Staff

Cuba, from one very difficult year to another

The very difficult year of 2025 is now behind us, and 2026 is slowly beginning. Despite the many shortages and difficulties, Cubans on the island once again put their resilience and willpower to the test, and tried to have the best possible New Year’s Eve, hoping for future improvement. But the change of year has not exactly brought good

Read at OnCuba News

arXiv (research papers) Hao Li SI

ReasAlign: Reasoning Enhanced Safety Alignment against Prompt Injection Attack

Large Language Models (LLMs) have enabled the development of powerful agentic systems capable of automating complex workflows across various fields. However, these systems are highly vulnerable to indirect prompt injection attacks, where malicious instructions embedded in external data can hijack agent behavior. In this work, we present ReasAlign, a model-level solution to improve safety alignment against indirect prompt injection attacks. The core idea of ReasAlign is to…

Read at arXiv (research papers)

Simon Willison SI

Claude Cowork Exfiltrates Files

Claude Cowork Exfiltrates Files Claude Cowork defaults to allowing outbound HTTP traffic to only a specific list of domains, to help protect the user against prompt injection attacks that exfiltrate their data. Prompt Armor found a creative workaround: Anthropic's API domain is on that list, so they constructed an attack that includes an attacker's own Anthropic API key and has the agent upload any files it can see to the https://api.anthropic.com/v1/files endpoint, allowing…

Read at Simon Willison

arXiv (research papers) Xinyi Wu SI

When Bots Take the Bait: Exposing and Mitigating the Emerging Social Engineering Attack in Web Automation Agent

Web agents, powered by large language models (LLMs), are increasingly deployed to automate complex web interactions. The rise of open-source frameworks (e.g., Browser Use, Skyvern-AI) has accelerated adoption, but also broadened the attack surface. While prior research has focused on model threats such as prompt injection and backdoors, the risks of social engineering remain largely unexplored. We present the first systematic study of social engineering attacks against web…

Read at arXiv (research papers)

arXiv (research papers) Shawn Li SI

Defenses Against Prompt Attacks Learn Surface Heuristics

Large language models (LLMs) are increasingly deployed in security-sensitive applications, where they must follow system- or developer-specified instructions that define the intended task behavior, while completing benign user requests. When adversarial instructions appear in user queries or externally retrieved content, models may override intended logic. Recent defenses rely on supervised fine-tuning with benign and malicious labels. Although these methods achieve high…

Read at arXiv (research papers)

The Guardian Communism Robert Tait in Washington Geopolitics

Trump’s Greenland threats echo dark moments of cold war alliances

Soviet invasions of allies helped destroy the Warsaw Pact – Trump’s dangerous rhetoric risks repeating the mistake inside Nato Donald Trump’s echoing of Russia’s talking points in its war against Ukraine has long been a cause for alarm and dismay in the west. Now an even more disturbing Kremlin precedent dating from the cold war is being evoked by the US president’s fixation on taking over Greenland – that of carrying out attacks on military allies.

Read at The Guardian Communism

Latinoamérica21 Ottón Solís Geopolitics

Without rules, the rule of the global gendarme

Beneath yesterday’s moral rhetoric and Trump’s barefaced cynicism today, the powers once again lay bare an uncomfortable truth: without rules or disguises, the United States presents itself as a global gendarme in the service of its own interests. O post Without rules, the rule of the global gendarme apareceu primeiro em Latinoamérica 21.

Read at Latinoamérica21

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)

arXiv (research papers) Alexandre Le Mercier SI

Hidden State Poisoning Attacks against Mamba-based Language Models

State space models (SSMs) like Mamba offer efficient alternatives to Transformer-based language models, with linear time complexity. Yet, their adversarial robustness remains critically unexplored. This paper studies the phenomenon whereby specific short input phrases induce a partial amnesia effect in such models, by irreversibly overwriting information in their hidden states, referred to as a Hidden State Poisoning Attack (HiSPA). Our benchmark RoBench-25 allows evaluating…

Read at arXiv (research papers)

Journal articles (OpenAlex) Jingfei Li · Open Journal of Social Sciences Geopolitics

A Study on the Pattern of China-U.S. Diplomatic Interaction Driven by Technological Competition —A Case Study of U.S. Export Controls on High-End Chips to China

Against the backdrop of the digital age, technological competition has become an important factor shaping great-power relations. U.S. export controls on high-end chips to China have pushed technology issues from the economic domain into the domain of strategic competition, thereby transforming the mode of China-U.S. diplomatic interaction. This paper argues that current China-U.S. diplomatic interaction exhibits the coexistence of strategic competition, policy contestation…

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Sarah Bufkin, Ida Danewid · South Atlantic Quarterly Geopolitics

Against Fascist Fantasies: Theorizing Racial Capitalism in (Post)Imperial Europe

This issue examines the history and contemporary articulations of racial capitalism within Europe. Challenging the idea that Europe is a homogeneous white entity, it interrogates how racialization and colonization have long been central to capital accumulation within the heart of empire. Today Europe's (post)imperial present remains underwritten by neocolonial relations of extraction, projects of Indigenous dispossession, and racialized regimes for controlling the mobility…

Read at Journal articles (OpenAlex)

arXiv (research papers) Yihan Wang SI

AegisAgent: An Autonomous Defense Agent Against Prompt Injection Attacks in LLM-HARs

The integration of Large Language Models (LLMs) into wearable sensing is creating a new class of mobile applications capable of nuanced human activity understanding. However, the reliability of these systems is critically undermined by their vulnerability to prompt injection attacks, where attackers deliberately input deceptive instructions into LLMs. Traditional defenses, based on static filters and rigid rules, are insufficient to address the semantic complexity of these…

Read at arXiv (research papers)

arXiv (research papers) Jaykumar Kasundra SI

AprielGuard

Safeguarding large language models (LLMs) against unsafe or adversarial behavior is critical as they are increasingly deployed in conversational and agentic settings. Existing moderation tools often treat safety risks (e.g. toxicity, bias) and adversarial threats (e.g. prompt injections, jailbreaks) as separate problems, limiting their robustness and generalizability. We introduce AprielGuard, an 8B parameter safeguard model that unify these dimensions within a single…

Read at arXiv (research papers)

Atlantic Council zbaddorf Geopolitics

Atlantic Council’s Strategic Litigation Project supports landmark criminal complaint in Argentina seeking accountability for crimes against humanity during Iran’s “Woman, Life, Freedom” Protests

BUENOS AIRES / WASHINGTON, DC — December 22, 2025 —The Atlantic Council’s Strategic Litigation Project supported the filing of the first-ever criminal complaint targeting Iranian officials for crimes against humanity during the “Woman, Life, Freedom” protest movement.

Read at Atlantic Council

Journal articles (OpenAlex) Michael Scroggins · Economic Anthropology Geopolitics

An Immodest Rejoinder: Illiberalism and Postliberalism, But Still Neoliberalism

ABSTRACT This article responds to the claims of Beck and High that the age of neoliberalism has passed and a new age of illiberalism is upon us. Against Beck and High, I argue that illiberalism is not opposed to neoliberalism, but rather accelerates it. In closing, I offer several practical strategies for navigating (thriving, even) in this illiberal, but still neoliberal, political moment.

Read at Journal articles (OpenAlex)

Journal articles (OpenAlex) Yannuo Xu · Interdisciplinary Humanities and Communication Studies Geopolitics

Nuclear Deterrence and International Law: Assessing Compliance with Article 2(4) of the UN Charter

This essay examines the legality and ethical implications of nuclear deterrence under international law, with a specific focus on Article 2(4) of the UN Charter, which prohibits the threat or use of force against the territorial integrity or political independence of states. It argues that nuclear deterrence, by its very nature, constitutes an unlawful threat of force and contravenes the foundational principles of international humanitarian law (IHL), including distinction…

Read at Journal articles (OpenAlex)

arXiv (research papers) Safwan Shaheer SI

Beyond the Benchmark: Innovative Defenses Against Prompt Injection Attacks

In this fast-evolving area of LLMs, our paper discusses the significant security risk presented by prompt injection attacks. It focuses on small open-sourced models, specifically the LLaMA family of models. We introduce novel defense mechanisms capable of generating automatic defenses and systematically evaluate said generated defenses against a comprehensive set of benchmarked attacks. Thus, we empirically demonstrated the improvement proposed by our approach in mitigating…

Read at arXiv (research papers)

Journal articles (OpenAlex) Steffen Köhn, Nestor Siré · Economic Anthropology Geopolitics

Virtual Pets, Real Precarity: Crypto Gaming as Speculative Labor in Inflation‐Ridden Cuba

ABSTRACT This article examines how Axie Infinity , a play‐to‐earn crypto game, became a precarious source of income for young Cubans amid an ongoing economic crisis. Against the backdrop of the COVID‐19 pandemic, intensified US sanctions, and a destabilizing currency reform, Cuba has faced severe shortages, runaway inflation, and growing social unrest. In this context, Axie Infinity offered an unexpected digital lifeline: Players could earn income by battling and trading…

Read at Journal articles (OpenAlex)

arXiv (research papers) Safwan Shaheer SI

Detecting Prompt Injection Attacks Against Application Using Classifiers

Prompt injection attacks can compromise the security and stability of critical systems, from infrastructure to large web applications. This work curates and augments a prompt injection dataset based on the HackAPrompt Playground Submissions corpus and trains several classifiers, including LSTM, feed forward neural networks, Random Forest, and Naive Bayes, to detect malicious prompts in LLM integrated web applications. The proposed approach improves prompt injection detection…

Read at arXiv (research papers)

Inter-American Dialogue Juanita Caballero Geopolitics

Chavez: “One of my hopes is that [Machado’s Nobel Peace Prize] gets the international community’s attention again to Venezuela, because ultimately any solution is going to require the support of the international community”

On December 12, President and CEO of the Inter-American Dialogue Rebecca Bill Chávez spoke with BBC News, where she discussed the consequences of seizing U.S. oil tankers off the Venezuelan coast.

Read at Inter-American Dialogue

arXiv (research papers) Najmul Hasan SI

Phishing Email Detection Using Large Language Models

Email phishing is one of the most prevalent and globally consequential vectors of cyber intrusion. As systems increasingly deploy Large Language Models (LLMs) applications, these systems face evolving phishing email threats that exploit their fundamental architectures. Current LLMs require substantial hardening before deployment in email security systems, particularly against coordinated multi-vector attacks that exploit architectural vulnerabilities. This paper proposes…

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)

arXiv (research papers) Zhibo Liang SI

SIEVE: Selective Integrity Verification and Escalation for Defending LLM Agents against Indirect Prompt Injection

Large Language Models (LLMs) are increasingly used as the core of agentic systems due to their strong reasoning, planning, and tool-use capabilities. By interacting with external environments, LLM agents can execute real-world tasks on behalf of users rather than merely generate text. This expanded capability also amplifies the threat of indirect prompt injection (IPI), where malicious external content can manipulate agent behavior and trigger unauthorized actions, privacy…

Read at arXiv (research papers)

Older stories