The security of LLM agents is critically dependent on their system prompt configuration, which creates a brittle attack surface that can be exploited by attackers inverting the prompt's core assumptio…

The paper proposes a lightweight hybrid MLP framework that uses structural URL features to achieve highly accurate and computationally efficient real-time phishing URL detection, outperforming several…

SnapGuard proposes a lightweight, multimodal method to detect prompt injection attacks in screenshot-based web agents by analyzing visual stability and contrast-polarity textual signals, achieving hig…

This paper proposes the first web-focused threat model for agentic browsers, demonstrating that traditional web social engineering attacks can be amplified into dangerous, reproducible threats when ex…

The paper demonstrates that autonomous web agents are highly susceptible to social-engineering attacks, leaking critical PII even when they internally flag a site as suspicious, necessitating output-l…

The paper demonstrates that autonomous web agents are highly susceptible to social-engineering attacks, leaking critical PII even when they internally flag a site as suspicious, necessitating output-l…

This paper analyzes various attack vectors against FIDO2 passkeys, demonstrating that while sophisticated attacks are possible, the overall security posture significantly raises the bar compared to tr…

WebTrap introduces a stealthy, mid-task hijacking attack that successfully compromises browser agents during long-horizon tasks by seamlessly fusing malicious instructions with the original user goal.

This paper provides a comprehensive, structured list of 42 email-based deception techniques, complete with 64 concrete examples, to serve as a modular reference for developing countermeasures.

The paper proposes WARD, a robust and efficient defense model that secures web agents against prompt injection attacks embedded in web content, achieving high recall and low false positives even again…

The paper proposes an on-device framework to detect and prevent the forwarding of images that have been physically recaptured (photographed) from a mobile screen, addressing the Screen Recaptured Anal…

The paper demonstrates a class of steganographic exfiltration attacks against vector databases by hiding data within embeddings, and proposes VectorPin, a cryptographic provenance protocol to detect s…

This paper introduces a systematic, privacy-preserving method using keystroke dynamics to robustly distinguish between human typing and automated HID injection attacks, independent of user identity.

This paper investigates a novel security vulnerability where imperceptible branding hints can be injected into images and subsequently re-rendered onto new objects by generative AI models, proposing b…

The paper introduces SET, a robust input-level backdoor detection framework that detects hidden malicious triggers in text-to-image diffusion models by analyzing systematic differences in how benign a…

The paper introduces VIPER, a novel backdoor attack framework that exploits the functional fusion of malicious and benign logic within dynamic prompt architectures, demonstrating a new, high-risk thre…

TraceScope is an interactive, sandboxed triage pipeline that analyzes complex phishing URLs by simulating human interaction and verifying suspicious behavior against a detailed checklist, achieving hi…

The paper proposes a unified, architecture-agnostic framework that significantly improves the robustness of deepfake image detectors against adversarial attacks by focusing on higher-order frequency s…

The paper proposes a lightweight, passive bot detection system using user-agent and favicon analysis on web server logs, achieving 67.7% bot detection with a low 3% false-positive rate.

The paper proposes a simple, generic attack strategy—re-watermarking—that reliably suppresses existing watermarks, demonstrating that watermarks can be used to attack other watermarks.