Protect AI Threat Research
# Vulnhuntr
# Remote Code Execution
# LLMs
Vulnhuntr: Autonomous AI Finds First 0-Day Vulnerabilities in Wild
Today, we introduce Vulnhuntr, a Python static code analyzer that leverages the power of large language models (LLMs) to find and explain complex, multistep vulnerabilities.
This report contains 32 vulnerabilities include a Denial of Service (DOS) in mintplex-labs/anything-llm and a Remote Code Execution (RCE) in LoLLMs.
# AI/ML Security
# Remote Code Execution
This report contains 31 vulnerabilities, including some critical vulnerabilities found in the Triton Inference Server and the Intel Neural Compressor.
# AI/ML Security
This report contains 20 vulnerabilities all found by our huntr community. Including some heavy hitters like an Local File Inclusion (LFI) in lollms and a Path Traversal vuln in AnythingLLM.
# LLMs
# Path Traversal
This report contains 20 vulnerabilities all discovered by our huntr community, including Authorization Bypass vuln in Lunary and a Server-Side Request Forgery (SSRF) vuln in Netaddr.
# Server-Side Request Forgery
This report contains 20 vulnerabilities found by the huntr community.
# AI/ML Security
# Remote Code Execution
# Server-Side Request Forgery