Remote Code Execution Vulnerability in Llama Stack by Facebook
CVE-2024-50050
Key Information:
- Vendor
Facebook
- Status
- Vendor
- CVE Published:
- 23 October 2024
Badges
What is CVE-2024-50050?
CVE-2024-50050 is a remote code execution vulnerability identified in the Llama Stack, a software stack developed by Facebook that facilitates specific network configurations and applications. This vulnerability originates from the use of the Python pickle
module for serialization in socket communications, which creates a significant security risk by allowing potential attackers to execute arbitrary code on affected systems. This can lead to severe security breaches and unauthorized access to sensitive data or system functionalities, posing a serious threat to organizations relying on this technology for their operations.
Technical Details
The vulnerability exists in versions of the Llama Stack prior to revision 7a8aa775e5a267cf8660d83140011a0b7f91e005. The core issue stems from the serialization format used for socket communication. The previous use of pickle
, a Python-specific serialization tool, is inherently insecure, as it can execute arbitrary code during de-serialization. Facebook has since updated the Llama Stack to use JSON for socket communication, which mitigates this risk significantly.
Potential impact of CVE-2024-50050
-
Unauthorized Remote Code Execution: Exploitation of this vulnerability can allow attackers to execute arbitrary code on the server, leading to complete control over the system, data manipulation, or installation of malicious software.
-
Data Breach Risks: Successful exploitation could lead to unauthorized access to sensitive organizational data, including personal information of users, proprietary business data, and other critical information, raising compliance and legal implications.
-
Operational Disruption: An attacker exploiting this vulnerability could disrupt normal operations by compromising applications built on the Llama Stack, potentially leading to service downtime, data loss, and a decrease in user trust.
News Articles

Critical Vulnerability in Meta Llama Framework Let Remote Attackers Execute Arbitrary Code
The Oligo Research team has disclosed a critical vulnerability in Meta’s widely used Llama-stack framework.

Critical Vulnerability in Meta's Llama Framework Exposes AI Systems to Remote Attacks
A critical security flaw, CVE-2024-50050, has been discovered in Meta’s Llama Stack framework, a widely used open-source tool for building and deploying generative AI (GenAI) applications.
References
Timeline
- 👾
Exploit known to exist
- 📰
First article discovered by The Hacker News
Vulnerability published