GuardReasoner-VL

GuardReasoner-VL：通过强化推理保护 VLM Yue Liu, Shengfang Zhai, Mingzhe Du Yulin Chen, Tri Cao, Hongcheng Gao, Cheng Wang Xinfeng Li, Kun Wang, Junfeng Fang, Jiaheng Zhang, Bryan Hooi 1新加坡国立大学, 2南洋理工大学 为了提升 VLM 的安全性，本文提出了一种新颖的基于推理的 VLM 护栏模型，命名为 GuardReasoner-VL。 其核心思想是通过在线 RL，激励护栏模型在做出审核决策之前进行深思熟虑的推理。...

langchain-nvidia-ai-endpoints has local file disclosure through VLM image inputs

Summary langchain-nvidia-ai-endpoints versions before 1.4.2 accepted local filesystem paths as image inputs for Vision Language Model VLM requests. If an application passed attacker-controlled image input to ChatNVIDIA or VLM reranking APIs, an attacker could cause files readable by the applicati...

GHSA-G28H-2CMM-RJ9X langchain-nvidia-ai-endpoints has local file disclosure through VLM image inputs

Summary langchain-nvidia-ai-endpoints versions before 1.4.2 accepted local filesystem paths as image inputs for Vision Language Model VLM requests. If an application passed attacker-controlled image input to ChatNVIDIA or VLM reranking APIs, an attacker could cause files readable by the applicati...

PT-2026-105510

Summary langchain-nvidia-ai-endpoints versions before 1.4.2 accepted local filesystem paths as image inputs for Vision Language Model VLM requests. If an application passed attacker-controlled image input to ChatNVIDIA or VLM reranking APIs, an attacker could cause files readable by the applicati...

AgentHijack: Visual Patch Attacks on Multimodal Computer-Use Agents

This paper presents an end-to-end evaluation framework for image-triggered command injection against computer-use agents CUAs. The goal is to test whether a local visual patch can induce verifiable environmental consequences along the full chain of screenshot input, VLM generation, action parsing...

CAPTCHAs in the Agentic Era: Solvers That Learn from Every Encounter

Vision-language models VLMs can solve visual CAPTCHAs without task-specific training, but the agents built on them approach every challenge from scratch. For such an agent, the hundredth instance of a familiar puzzle costs as much time and compute as the first. Specialized detectors invert the...

Once Poisoned, Arbitrarily Controlled: A Programmable Backdoor in VLMs

Existing vision-language model VLM backdoors are usually treated as static vulnerabilities: one-to-one and N-to-N attacks bind one or more triggers to a finite set of targets before victim training. This assumption substantially underestimates the threat. We show that a single poisoning phase can...

PYSEC-2026-1965 Hugging Face Text Generation Inference vulnerable to Uncontrolled Resource Consumption

A vulnerability in huggingface/text-generation-inference version 3.3.6 allows unauthenticated remote attackers to exploit unbounded external image fetching during input validation in VLM mode. The issue arises when the router scans inputs for Markdown image links and performs a blocking HTTP GET...

CVE-2026-0599

A vulnerability in huggingface/text-generation-inference version 3.3.6 allows unauthenticated remote attackers to exploit unbounded external image fetching during input validation in VLM mode. The issue arises when the router scans inputs for Markdown image links and performs a blocking HTTP GET...

GHSA-J7X9-7J54-2V3H Hugging Face Text Generation Inference vulnerable to Uncontrolled Resource Consumption

A vulnerability in huggingface/text-generation-inference version 3.3.6 allows unauthenticated remote attackers to exploit unbounded external image fetching during input validation in VLM mode. The issue arises when the router scans inputs for Markdown image links and performs a blocking HTTP GET...

Hugging Face Text Generation Inference vulnerable to Uncontrolled Resource Consumption

A vulnerability in huggingface/text-generation-inference version 3.3.6 allows unauthenticated remote attackers to exploit unbounded external image fetching during input validation in VLM mode. The issue arises when the router scans inputs for Markdown image links and performs a blocking HTTP GET...

CVE-2026-0599: Uncontrolled Resource Consumption

A vulnerability in huggingface/text-generation-inference version 3.3.6 allows unauthenticated remote attackers to exploit unbounded external image fetching during input validation in VLM mode. The issue arises when the router scans inputs for Markdown image links and performs a blocking HTTP GET...

MAI-2024-0012

Large Vision-Language Models VLMs are susceptible to a sophisticated black-box jailbreak attack known as IDEATOR. This attack utilizes a separate VLM to craft malicious image-text pairs designed to circumvent the target VLM's safety protocols. By iteratively refining prompts based on the target...