An Empirical Evaluation of Prompt Injection Vulnerabilities in Large Language Models across Multilingual and Obfuscated Attack Scenarios

Large Language Models LLMs have rapidly evolved, transforming industries by automating complex tasks and generating human-like content. However, as their adoption accelerates, prompt injection vulnerabilities have become increasingly apparent. Malicious actors exploit these weaknesses to generate...

PT-2026-52680

Name of the Vulnerable Software and Affected Versions Claude Code versions 2.1.38 through 2.1.162 Description Worktree handling allows the creation of worktrees named ".git" and navigation to worktrees outside the sandbox context, enabling git directory confusion attacks. By exploiting symlink...

GO-2026-5631 k8sGPT has Prompt Injection through its k8sGPT-Operator in github.com/k8sgpt-ai/k8sgpt

k8sGPT has Prompt Injection through its k8sGPT-Operator in github.com/k8sgpt-ai/k8sgpt...

Malicious code in unsafe-malicious-package (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector 3579cb796e48f446b07e2dbbce2e301d1a3e87d8a9a35ed1dbe825fc53f29da9 On npm install, the package's postinstall lifecycle script scripts/postinstall.js reads the installer's AWS credentials file at /.aws/credentials and...

MAL-2026-6486 Malicious code in unsafe-malicious-package (npm)

--- -= Per source details. Do not edit below this line.=- Source: amazon-inspector 3579cb796e48f446b07e2dbbce2e301d1a3e87d8a9a35ed1dbe825fc53f29da9 On npm install, the package's postinstall lifecycle script scripts/postinstall.js reads the installer's AWS credentials file at /.aws/credentials and...

Interesting Paper Exploring Prompt Injection

This is a fascinating explotation of how LLMs fall for prompt injection attacks. It turns out that they learn to recognize the style of text in different role/instruction blocks, and not just the tags. Their conclusion: Role tags were a formatting trick that became the security architecture and t...

New Gaslight macOS Malware Uses Prompt Injection to Disrupt AI-Assisted Analysis

A previously undocumented Rust-based macOS implant and information stealer has been found to embed a prompt injection payload designed to trick a malware analyst's artificial intelligence AI tools and trick it into aborting or refusing an analysis of the artifact. The malware has been codenamed...

DRUPAL-CONTRIB-2026-054

The module and certain submodules AI Automators, AI Translate, AI API Explorer, AI Content Suggestions provide the ability to use an LLM to generate HTML or Markdown and preview it in a browser. Under certain circumstances, rendering of this HTML can lead to Cross Site Scripting, or exposing secr...

FlowiseAI Flowise CSV Agent Prompt Injection Remote Code Execution Vulnerability

This vulnerability allows remote attackers to execute arbitrary code on affected installations of Flowise. Authentication is not required to exploit this vulnerability. The specific flaw exists within the run method of the CSVAgents class. The issue results from insufficient input sanitization wh...

PT-2026-52167

Name of the Vulnerable Software and Affected Versions Drupal AI versions 0.0.0 through 1.2.17 Drupal AI versions 1.3.0 through 1.3.8 Drupal AI versions 1.4.0 through 1.4.3 Description Improper neutralization of input during web page generation allows Cross-Site Scripting XSS. The module and...

AI (Artificial Intelligence) - Moderately critical - Information Disclosure / Cross-site Scripting - SA-CONTRIB-2026-054

The module and certain submodules AI Automators, AI Translate, AI API Explorer, AI Content Suggestions provide the ability to use an LLM to generate HTML or Markdown and preview it in a browser. Under certain circumstances, rendering of this HTML can lead to Cross Site Scripting, or exposing secr...

CVE-2026-54009

Open WebUI is a self-hosted artificial intelligence platform designed to operate entirely offline. Prior to 0.9.6, POST /api/chat/completions accepts an imageurl.url value that, when it does NOT start with http://, https://, or data:image/, is interpreted as a file id and resolved against the...

CVE-2026-56696 OpenHarness - Prompt Injection via /issue and /pr_comments Slash Commands

OpenHarness /issue and /prcomments slash commands lack remoteinvocable=False protection, allowing remote channel senders to write attacker-controlled Markdown into project context files. Admitted remote attackers can inject malicious content into .openharness/issue.md and .openharness/prcomments....

CVE-2026-56696 OpenHarness - Prompt Injection via /issue and /pr_comments Slash Commands

OpenHarness /issue and /prcomments slash commands lack remoteinvocable=False protection, allowing remote channel senders to write attacker-controlled Markdown into project context files. Admitted remote attackers can inject malicious content into .openharness/issue.md and .openharness/prcomments....

CVE-2026-56696

CVE-2026-56696 affects OpenHarness; the /issue and /pr_comments slash commands lack remote_invocable=False protection. This allows remote attackers to write attacker-controlled Markdown into project context files (.openharness/issue.md and .openharness/pr_comments.md). The injected content is sub...

CVE-2026-56696: Missing Authorization

OpenHarness /issue and /prcomments slash commands lack remoteinvocable=False protection, allowing remote channel senders to write attacker-controlled Markdown into project context files. Admitted remote attackers can inject malicious content into .openharness/issue.md and .openharness/prcomments....

CVE-2025-66389

GitHub Copilot 1.372.0 allows filesystem access outside of a workspace folder without user approval via a file-handler URI parameter to fetchwebpage. Therefore, exfiltration could occur if there is indirect prompt injection...

CVE-2025-66389

GitHub Copilot version 1.372.0 is vulnerable to unauthorized filesystem access outside of a workspace folder. The root cause is a file-handler URI parameter used in the fetch_webpage() function, which allows access without user approval. This flaw could lead to data exfiltration if an indirect pr...

CVE-2025-66389

GitHub Copilot 1.372.0 allows filesystem access outside of a workspace folder without user approval via a file-handler URI parameter to fetchwebpage. Therefore, exfiltration could occur if there is indirect prompt injection...

EUVD-2025-210298

GitHub Copilot 1.372.0 allows filesystem access outside of a workspace folder without user approval via a file-handler URI parameter to fetchwebpage. Therefore, exfiltration could occur if there is indirect prompt injection...