CVE-2026-4503

IBM Langflow Desktop 1.0.0 through 1.8.4 Langflow could allow an unauthenticated user to view other users' images due to an indirect object reference through a user-controlled key...

CVE-2026-4503 Unauthenticated Insecure Direct Object Reference (IDOR) Vulnerability in Langflow Desktop Image Download Endpoint

IBM Langflow Desktop 1.0.0 through 1.8.4 Langflow could allow an unauthenticated user to view other users' images due to an indirect object reference through a user-controlled key...

EUVD-2026-26435

IBM Langflow Desktop 1.0.0 through 1.8.4 Langflow could allow an unauthenticated user to view other users' images due to an indirect object reference through a user-controlled key...

IBM Langflow Desktop 安全漏洞

IBM Langflow Desktop is a desktop application for AI process orchestration developed by IBM. Versions 1.0.0 to 1.8.4 of IBM Langflow Desktop contain security vulnerabilities. These vulnerabilities stem from indirect object references using user control keys, which may allow unauthenticated users ...

PT-2026-36190

Name of the Vulnerable Software and Affected Versions IBM Langflow Desktop versions 1.0.0 through 1.8.4 Description An unauthenticated user can view images belonging to other users. This is possible due to an indirect object reference through a user-controlled key. Recommendations At the moment,...

Indirect Prompt Injection in the Wild: An Empirical Study of Prevalence, Techniques, and Objectives

As LLMs are increasingly integrated into systems that browse, retrieve, summarize, and act on web content, webpages have become an untrusted input vector for downstream model behavior. This enables site owners, contributors, and adversaries to embed instructions directly in web resources, i.e.,...

Logic-to-Code Execution via Indirect Prompt Injection

This document explores a critical architectural vulnerability in Large Language Model LLM implementations, specifically within Command Line Interface CLI tools and automated agentic workflows. The research demonstrates how the absence of separation between the control plane instructions and the...

CVE-2026-41277 Flowise: Mass Assignment in DocumentStore Create Endpoint Leads to Cross-Workspace Object Takeover (IDOR)

Flowise is a drag & drop user interface to build a customized large language model flow. Prior to 3.1.0, a Mass Assignment vulnerability in the DocumentStore creation endpoint allows authenticated users to control the primary key id and internal state fields of DocumentStore entities. Because the...

Hackers Use Hidden Website Instructions in New Attacks on AI Assistants

Cybersecurity researchers at Forcepoint uncover new indirect prompt injection attacks that use hidden website code to exploit AI assistants like GitHub Copilot...

CVE-2026-40907 WWBN AVideo has IDOR in Live Restreams list.json.php that Exposes Other Users' Stream Keys and OAuth Tokens

WWBN AVideo is an open source video platform. In versions 29.0 and prior, the endpoint plugin/Live/view/Liverestreams/list.json.php contains an Insecure Direct Object Reference IDOR vulnerability that allows any authenticated user with streaming permission to retrieve other users' live restream...

Unity Linux 20.1050e / 20.1070e Security Update: kernel (UTSA-2026-010885)

The Unity Linux 20 host has a package installed that is affected by a vulnerability as referenced in the UTSA-2026-010885 advisory. In the Linux kernel, the following vulnerability has been resolved: ext4: add idatasem protection in ext4destroyinlinedatanolock Fix a race between inline data...

📄 WBCE CMS Privilege Escalation / Insecure Direct Object Reference

WBCE CMS versions prior to 1.6.4 suffers from insecure direct object reference and privilege escalation vulnerabilities. CVE-2025-65094: WBCE CMS is Vulnerable to Privilege Escalation via Group ID Manipulation IDOR Overview | Field | Details | |---|---| | CVE ID | CVE-2025-65094 | | Severity | HI...

CVE-2026-39853

osslsigncode is a tool that implements Authenticode signing and timestamping. Prior to 2.12, A stack buffer overflow vulnerability exists in osslsigncode in several signature verification paths. During verification of a PKCS7 signature, the code copies the digest value from a parsed...

CVE-2026-39853

osslsigncode is a tool that implements Authenticode signing and timestamping. Prior to 2.12, A stack buffer overflow vulnerability exists in osslsigncode in several signature verification paths. During verification of a PKCS7 signature, the code copies the digest value from a parsed...

EUVD-2026-20942

osslsigncode is a tool that implements Authenticode signing and timestamping. Prior to 2.12, A stack buffer overflow vulnerability exists in osslsigncode in several signature verification paths. During verification of a PKCS7 signature, the code copies the digest value from a parsed...

CVE-2026-39853 osslsigncode has a Stack Buffer Overflow via Unbounded Digest Copy During Signature Verification

osslsigncode is a tool that implements Authenticode signing and timestamping. Prior to 2.12, A stack buffer overflow vulnerability exists in osslsigncode in several signature verification paths. During verification of a PKCS7 signature, the code copies the digest value from a parsed...

@0xwork/connect (>=0.1.0 <=0.1.7), @agentholdings/agent-passport (>=0.1.0 <=0.1.5) +23 more potentially affected by CVE-2026-40037 via openclaw (>=2026.3.22 <=2026.4.5)

openclaw NPM version =2026.3.22, =0.1.0, =0.1.0, =0.8.3, =0.1.0, =2026.3.25, =0.0.0, =27.2.5, =1.1.0, =2.1.3, =2026.3.24-3, =0.14.39, =0.1.0, =0.1.1, =0.2.18 and more Source cves: CVE-2026-40037 Source advisory: SNYK:JS-OPENCLAW-15949297...

GrafanaGhost Vulnerability Allows Data Theft via AI Injection

GrafanaGhost is a critical vulnerability in Grafana’s AI components that uses indirect prompt injection and protocol-relative URL bypasses to exfiltrate data...

Your Agent Is More Brittle Than You Think: Uncovering Indirect Injection Vulnerabilities in Agentic LLMs

The rapid deployment of open-source frameworks has significantly advanced the development of modern multi-agent systems. However, expanded action spaces, including uncontrolled privilege exposure and hidden inter-system interactions, pose severe security challenges. Specifically, Indirect Prompt...

Architecting Secure AI Agents: Perspectives on System-Level Defenses against Indirect Prompt Injection Attacks

AI agents, predominantly powered by large language models LLMs, are vulnerable to indirect prompt injection, in which malicious instructions embedded in untrusted data can trigger dangerous agent actions. This position paper discusses our vision for system-level defenses against indirect prompt...