arXiv subjects

Wonwoo Choi

Publications and source records attributed to Wonwoo Choi.

4 recordsLinked to original sources

arXiv2026

MEMORY Wins All: Indirect Bias Injection Attacks via Social Media Feeds

Personal AI agents routinely consume external content while performing tasks such as web browsing, email processing, and SNS feed summarization, and they retain selected information or execution results in persistent memory for later use. We show that this ordinary ingestion of external content opens an indirect path for manipulating subsequent agent behavior. Based on this observation, we present IBIA, an Indirect Bias Injection Attack that plants an adversary-aligned stance on a specific topic into a victim agent's memory through external content, without direct access to the agent, its memory, or future user queries. For this, IBIA combines three mechanisms: comment cloaking, which keeps the crafted content consistent with the surrounding discussion, comment watermarking, which enables lightweight identification during curation, and category anchoring, which makes the retained stance salient under later related requests. We evaluate IBIA on BiasBench, a benchmark of 6,000 adversary-crafted social comments and 120 email instances. The watermark-based curation identifies 95.9% of the injected comments. Under the OpenClaw setting, IBIA achieves adversary-aligned response rates (AARs) of 91.2% on average across four downstream tasks, including 86.6% on the frontier GPT-5.5. We further propose a memory boundary defense that detects the injected bias and reduces AARs to 80.6%.

cs.AI↗

arXiv2026

$PC^2$: Politically Controversial Content Generation via Jailbreaking Attacks on GPT-based Text-to-Image Models

The rapid evolution of text-to-image (T2I) models has enabled high-fidelity visual synthesis on a global scale. However, these advancements have introduced significant security risks, particularly regarding the generation of harmful content. Politically harmful content, such as fabricated depictions of public figures, poses severe threats when weaponized for fake news or propaganda. Despite its criticality, the robustness of current T2I safety filters against such politically motivated adversarial prompting remains underexplored. In response, we propose $PC^2$, the first black-box political jailbreaking framework for T2I models. It exploits a novel vulnerability where safety filters evaluate political sensitivity based on linguistic context. $PC^2$ operates through: (1) Identity-Preserving Descriptive Mapping to obfuscate sensitive keywords into neutral descriptions, and (2) Geopolitically Distal Translation to map these descriptions into fragmented, low-sensitivity languages. This strategy prevents filters from constructing toxic relationships between political entities within prompts, effectively bypassing detection. We construct a benchmark of 240 politically sensitive prompts involving 36 public figures. Evaluation on commercial T2I models, specifically the GPT series, shows that while all original prompts are blocked, $PC^2$ achieves attack success rates (ASRs) of up to 86% and outperforms state-of-the-art frameworks by a large margin. We further propose a ready-to-deploy multi-layered filtering mitigation against $PC^2$-style attacks, reducing ASR to approximately 10%.

cs.CR↗

arXiv2025

AutoPatch: Multi-Agent Framework for Patching Real-World CVE Vulnerabilities

Large Language Models (LLMs) have emerged as promising tools in software development, enabling automated code generation and analysis. However, their knowledge is limited to a fixed cutoff date, making them prone to generating code vulnerable to newly disclosed CVEs. Frequent fine-tuning with new CVE sets is costly, and existing LLM-based approaches focus on oversimplified CWE examples and require providing explicit bug locations to LLMs, limiting their ability to patch complex real-world vulnerabilities. To address these limitations, we propose AutoPatch, a multi-agent framework designed to patch vulnerable LLM-generated code, particularly those introduced after the LLMs' knowledge cutoff. AutoPatch integrates Retrieval-Augmented Generation (RAG) with a structured database of recently disclosed vulnerabilities, comprising 525 code snippets derived from 75 high-severity CVEs across real-world systems such as the Linux kernel and Chrome. AutoPatch combines semantic and taint analysis to identify the most relevant CVE and leverages enhanced Chain-of-Thought (CoT) reasoning to construct enriched prompts for verification and patching. Our unified similarity model, which selects the most relevant vulnerabilities, achieves 90.4 percent accuracy in CVE matching. AutoPatch attains 89.5 percent F1-score for vulnerability verification and 95.0 percent accuracy in patching, while being over 50x more cost-efficient than traditional fine-tuning approaches.

cs.CR↗

arXiv2017

Short-Channel Field Effect Transistors with 9-Atom and 13-Atom wide Graphene Nanoribbons

Bottom-up synthesized GNRs and GNR heterostructures have promising electronic properties for high performance field effect transistors (FETs) and ultra-low power devices such as tunnelling FETs. However, the short length and wide band gap of these GNRs have prevented the fabrication of devices with the desired performance and switching behaviour. Here, by fabricating short channel (Lch ~20 nm) devices with a thin, high-k gate dielectric and a 9-atom wide (0.95 nm) armchair GNR as the channel material, we demonstrate FETs with high on-current (Ion >1 uA at Vd = -1 V) and high Ion/Ioff ~10^5 at room temperature. We find that the performance of these devices is limited by tunnelling through the Schottky barrier (SB) at the contacts and we observe an increase in the transparency of the barrier by increasing the gate field near the contacts. Our results thus demonstrate successful fabrication of high performance short-channel FETs with bottom-up synthesized armchair GNRs.

cond-mat.mes-hall↗