This blog details research and analysis of an active campaign that exploits a critical unauthenticated remote code execution (RCE) vulnerability, CVE-2025-3248, that has been identified in Langflow versions prior to 1.3.0. Langflow is a Python-powered visual framework for building AI applications with over 70,000 GitHub stars, and its versions prior to 1.3.0 contains a flaw in its code validation mechanism that permits arbitrary code execution.
Unauthenticated attackers can exploit this vulnerability by crafting malicious POST requests to the /api/v1/validate/code endpoint. The malicious payload in our investigation was found embedded within argument defaults or decorators of a Python function definition. Since Langflow does not enforce input validation or sandboxing, these payloads are compiled and executed within the server’s context, leading to RCE.
Read more…
Source: Trend Micro
Sign up for our Newsletter
The latest news and insights delivered right to your inbox.
Related:
- Dridex Campaigns Hitting Millions of Recipients Using Unpatched Microsoft Zero-Day
April 10, 2017
This weekend saw multiple reports of a new zero-day vulnerability that affected all versions of Microsoft Word. Today, Proofpoint researchers observed the document exploit being used in a large email campaign distributing the Dridex banking Trojan. This campaign was sent to millions of recipients across numerous organizations primarily in Australia. This represents a significant level of ...
- Apple Fixes 223 Vulnerabilities Across macOS, iOS, Safari
March 28, 2017
Apple fixed hundreds of bugs, 223 to be exact, across a slate of products including macOS Sierra, iOS, Safari, watchOS, and tvOS on Monday. More than a quarter of the bugs, 40 in macOS Sierra, and 30 in iOS, could lead to arbitrary code execution – in some instances with root privileges, Apple warned. The lion’s share of ...
- Cisco Warns of Critical Vulnerability Revealed in ‘Vault 7’ Data Dump
March 20, 2017
Cisco Systems warned customers on Friday of a critical vulnerability that could allow an attacker to execute arbitrary code and obtain full control on more than 300 different models of its switches and routers. Cisco said it became aware of the vulnerability after WikiLeaks released its Vault 7 cache of documents that revealed the existence ...