CVE-2024-50050

Published Oct 23, 2024

Last updated a year ago

Overview

Description
Llama Stack prior to revision 7a8aa775e5a267cf8660d83140011a0b7f91e005 used pickle as a serialization format for socket communication, potentially allowing for remote code execution. Socket communication has been changed to use JSON instead.
Source
cve-assign@fb.com
NVD status
Awaiting Analysis

Risk scores

CVSS 3.1

Type
Secondary
Base score
6.3
Impact score
3.4
Exploitability score
2.8
Vector string
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:L/I:L/A:L
Severity
MEDIUM

Social media

Hype score
Not currently trending
  1. https://t.co/oWC03dkQfK ローカルLLMにも脆弱性があり、「自分だけで使っているから安全」は誤解だと解説します。 CVE-2024-50050等、ネットワーク経由や悪意あるモデルファイルによる攻撃リスクを紹介。 利用形態ご

    @topickapp_com

    2 Mar 2026

    63 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  2. ローカルLLMの脆弱性があるんか?〜CVE-2024-50050から学ぶ「自分だけで使ってるから安全」の落とし穴〜 - Qiita https://t.co/7TJ7l3DOSN

    @NEWS_0

    1 Mar 2026

    196 Impressions

    0 Retweets

    0 Likes

    1 Bookmark

    0 Replies

    0 Quotes

  3. #ローカルLLM にも #セキュリティーホール があるというお話 ローカルLLMの脆弱性があるんか?〜CVE-2024-50050から学ぶ「自分だけで使ってるから安全」の落とし穴〜 #Python - Qiita https://t.co/dRNbKwEttn https://t.co/WyHANT

    @hosikuzunoie

    1 Mar 2026

    80 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  4. 【読んだ】ローカルLLMの脆弱性があるんか?〜CVE-2024-50050から学ぶ「自分だけで使ってるから安全」の落とし穴〜 - Qiita ( https://t.co/XSZqncOBej )

    @nonokao

    1 Mar 2026

    139 Impressions

    0 Retweets

    1 Like

    0 Bookmarks

    0 Replies

    0 Quotes

  5. ローカルLLMの脆弱性があるんか?〜CVE-2024-50050から学ぶ「自分だけで使ってるから安全」の落とし穴〜 #Python - Qiita https://t.co/aBfvKHomHl

    @ohhara_shiojiri

    1 Mar 2026

    83 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  6. ローカルLLMの脆弱性があるんか?〜CVE-2024-50050から学ぶ「自分だけで使ってるから安全」の落とし穴〜 - Qiita https://t.co/Jhhtz8zHJr

    @matsuu_zatsu

    1 Mar 2026

    239 Impressions

    2 Retweets

    5 Likes

    2 Bookmarks

    0 Replies

    0 Quotes

  7. > ローカルLLMの脆弱性があるんか?〜CVE-2024-50050から学ぶ「自分だけで使ってるから安全」の落とし穴〜 - Qiita 🅱 はてなブックマーク テクノロジー新着 より https://t.co/QtUmUMtGSW

    @mohritaroh

    28 Feb 2026

    101 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  8. ローカルLLMの脆弱性があるんか?〜CVE-2024-50050から学ぶ「自分だけで使ってるから安全」の落とし穴〜 #Python - Qiita https://t.co/X1u2XuGAs3 "この記事では、ローカルLLM界隈で実際に報告された脆弱性を整理し、「結

    @catnap707

    28 Feb 2026

    238 Impressions

    0 Retweets

    1 Like

    2 Bookmarks

    0 Replies

    0 Quotes

  9. [2026/03/01 06:00] トレンド4位 ローカルLLMの脆弱性があるんか?〜CVE-2024-50050から学ぶ「自分だけで使ってるから安全」の落とし穴〜 by GeneLab_999 https://t.co/ZGcA35T0yz

    @QiitaTrend

    28 Feb 2026

    106 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  10. [2026/02/28 18:00] トレンド4位 ローカルLLMの脆弱性があるんか?〜CVE-2024-50050から学ぶ「自分だけで使ってるから安全」の落とし穴〜 by GeneLab_999 https://t.co/ZGcA35T0yz

    @QiitaTrend

    28 Feb 2026

    78 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  11. ローカルLLMの脆弱性があるんか?〜CVE-2024-50050から学ぶ「自分だけで使ってるから安全」の落とし穴〜 https://t.co/bBndn5fmE7 #Qiita

    @qiitapoi

    27 Feb 2026

    430 Impressions

    0 Retweets

    2 Likes

    2 Bookmarks

    0 Replies

    0 Quotes

  12. ローカルLLMの脆弱性があるんか?〜CVE-2024-50050から学ぶ「自分だけで使ってるから安全」の落とし穴〜 #Python - Qiita https://t.co/s27GchLxy0

    @swtdeadlypoison

    27 Feb 2026

    50 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  13. #後で読む 用メモです→ ローカルLLMの脆弱性があるんかCVE-2024-50050から学ぶ自分だけで使ってるから安全の落とし穴 https://t.co/hlkmay2815

    @TommiyTw

    27 Feb 2026

    40 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  14. 【新着記事】ローカルLLMの重大な脆弱性!CVE-2024- #ローカルLLM #CVE-2024-50050 #セキュリティ #AI #LLM https://t.co/nsIcmBo6sT

    @warokai_blog

    26 Feb 2026

    67 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  15. New research: Dual Critical Failures in ChatGPT Canvas — Pickle RCE + DNS Exfiltration via browser rendering Same vuln class Meta patched in 11 days (CVE-2024-50050). OpenAI dismissed it. https://t.co/W2p9WrhwKh #infosec #AIsecurity #ChatGPT

    @SnailSploit

    1 Feb 2026

    0 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  16. 🔐 New research alert! #LlamaStack-RCE (CVE-2024-50050) demonstrates deterministic supply chain exploitation via PVM opcode manipulation & ZeroMQ injection. 🧠 A critical deep dive into #AISecurity & #SupplyChainPoisoning. https://t.co/0rS0HzZJmh

    @TheExploitLab

    19 Jan 2026

    64 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  17. CVE-2024-50050 infected Meta, NVIDIA, Microsoft, vLLM, SGLang—because they copy-pasted the same insecure code from each other. Two remain unpatched in Dec 2025. This is AI's "move fast" culture in action. https://t.co/ZvYhVubY2p #CyberSecurity #AI

    @Mohamed_Writes

    10 Dec 2025

    4 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  18. Imagine building a house of cards in a windstorm, but the cards are borrowed from your neighbor's deck, flaws and all. That's ShadowMQ in a nutshell, per Oligo's report. It kicked off with CVE-2024-50050 in Llama, where network data got deserialized without checks, opening ht

    @kenshii_ai

    15 Nov 2025

    42 Impressions

    1 Retweet

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  19. Self-hosted Llama 3 runs 9x faster than GPT-4o 🚀 But CVE-2024-50050 shows "free" has hidden security costs. Smart play: Start with APIs → migrate high-volume to self-hosted Your LLM strategy?

    @Mike_ChenWM

    6 Oct 2025

    10 Impressions

    0 Retweets

    1 Like

    0 Bookmarks

    0 Replies

    0 Quotes

  20. 🚨 Vulnerabilidad crítica en el Framework Llama de Meta CVE-2024-50050 expone los sistemas de Inteligencia Artificial a ataques de ejecución remota de código. Actualiza a la última versión para mantenerte seguro. 🔒 #Ciberseguridad #IA #Meta https://t.co/sf9IqhIzge #Llama4

    @henryraul

    26 Feb 2025

    185 Impressions

    8 Retweets

    11 Likes

    1 Bookmark

    0 Replies

    0 Quotes

  21. Actively exploited CVE : CVE-2024-50050

    @transilienceai

    23 Feb 2025

    16 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    1 Reply

    0 Quotes

  22. Actively exploited CVE : CVE-2024-50050

    @transilienceai

    22 Feb 2025

    12 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    1 Reply

    0 Quotes

  23. Actively exploited CVE : CVE-2024-50050

    @transilienceai

    22 Feb 2025

    11 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    1 Reply

    0 Quotes

  24. Actively exploited CVE : CVE-2024-50050

    @transilienceai

    19 Feb 2025

    4 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    1 Reply

    0 Quotes

  25. Vulnerabilidad framework Llama de Meta expone los sistemas de IA a riesgos de ejecución remota de código Meta's Llama large language model (LLM) CVE-2024-50050 https://t.co/5X7zHPsWOz… https://t.co/kx0lPAOysv

    @doncaptador

    17 Feb 2025

    15 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  26. Vulnerabilidad framework Llama de Meta expone los sistemas de IA a riesgos de ejecución remota de código Meta's Llama large language model (LLM) CVE-2024-50050 https://t.co/bW8bsXugHa https://t.co/spAvq3ROiT

    @elhackernet

    17 Feb 2025

    2577 Impressions

    11 Retweets

    27 Likes

    3 Bookmarks

    0 Replies

    0 Quotes

  27. Actively exploited CVE : CVE-2024-50050

    @transilienceai

    15 Feb 2025

    11 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    1 Reply

    0 Quotes

  28. Actively exploited CVE : CVE-2024-50050

    @transilienceai

    15 Feb 2025

    12 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    1 Reply

    0 Quotes

  29. Actively exploited CVE : CVE-2024-50050

    @transilienceai

    13 Feb 2025

    11 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    1 Reply

    0 Quotes

  30. Actively exploited CVE : CVE-2024-50050

    @transilienceai

    5 Feb 2025

    16 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    1 Reply

    0 Quotes

  31. برای فریمورک Llama-stack آسیب پذیری با کد شناسایی CVE-2024-50050 و از نوع RCE منتشر شده است. LIama-stack فریمورک ای است که از سال ۲۰۲۴ شروع به فعالیت نموده است و برای برنامه نویسی با AI نمره این آسیب پذیری 9.8 می باشد. https://t.co/Poz3aKY03t https://t.co/81ummVqLug

    @AmirHossein_sec

    30 Jan 2025

    28 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  32. Critical security flaw (CVE-2024-50050) found in Meta's Llama-Stack risks Generative AI apps. Affected by unsafe pyzmq use, patch released to upgrade to v0.0.41. ⚠️ #Meta #GenerativeAI #USA link: https://t.co/YRNWKZEWqS https://t.co/YbqNYNjGAc

    @TweetThreatNews

    27 Jan 2025

    30 Impressions

    0 Retweets

    1 Like

    0 Bookmarks

    0 Replies

    0 Quotes

  33. 🚨 Critical Vulnerabilities in Meta's Llama Framework and Other AI Systems Highlight Growing Cybersecurity Challenges 🚨 Meta Llama Framework Vulnerability: CVE-2024-50050 Severity: CVSS: 6.3 (Meta) Critical: 9.3 (Snyk) Issue: Arbitrary code execution via deserialization of… ht

    @GHak2learn27752

    27 Jan 2025

    82 Impressions

    0 Retweets

    2 Likes

    1 Bookmark

    1 Reply

    1 Quote

  34. The vulnerability, tracked as CVE-2024-50050, has been assigned a CVSS score of 6.3 out of 10.0. Supply chain security firm Snyk, on the other hand, has assigned it a critical severity rating of 9.3. https://t.co/oZw0a3Xs6z

    @jbhall56

    27 Jan 2025

    5 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  35. CVE-2024-50050 impacts Meta's Llama Stack framework #CVE-2024-50050 #LlamaFramework #META https://t.co/IuZh6WnvPp

    @pravin_karthik

    27 Jan 2025

    31 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  36. Cybersecurity continues to face challenges as AI technologies grow in complexity and integration. Recently, a critical security flaw (CVE-2024-50050) was identified in Meta’s Llama framework, a tool designed to support AI application development. Watch the video till the end to

    @FORTBRIDGE

    27 Jan 2025

    47 Impressions

    0 Retweets

    3 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  37. Top 5 Trending CVEs: 1 - CVE-2025-23006 2 - CVE-2024-50050 3 - CVE-2024-43468 4 - CVE-2025-0282 5 - CVE-2025-21298 #cve #cvetrends #cveshield #cybersecurity https://t.co/4Fua3CAN6W

    @CVEShield

    27 Jan 2025

    9 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  38. Se ha identificado una falla de alta gravedad en el framework Llama de Meta que permite la ejecución remota de código mediante la deserialización de datos no confiables. Esta vulnerabilidad, catalogada como CVE-2024-50050, podría ser utilizada para ejecutar código arbitrario en…

    @citarafy

    26 Jan 2025

    21 Impressions

    0 Retweets

    1 Like

    1 Bookmark

    0 Replies

    0 Quotes

  39. Vulnerabilidad critica en el stack de LLama de Meta. Esta vulnerabilidad que esta bajo el CVE-2024-50050, es una posible ejecución arbitraria de código en servidor. https://t.co/M9bZmpcrXi #vulnerability #LLM #hacking

    @revers3shell

    26 Jan 2025

    8 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    1 Reply

    0 Quotes

  40. CVE-2024-50050: vulnerabilità critica in Meta Llama Stack Sicurezza Informatica, aggiornamento, CVE-2024-50050, exploit, Intelligenza Artificiale, LLM, Meta Llama Stack, open-source, pyzmq pickle, sicurezza AI, vulnerabilità https://t.co/2WWyRf24o9 https://t.co/95LQ4ktmII

    @matricedigitale

    26 Jan 2025

    31 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

  41. 1/ 🚨 메타의 라마(LLaMA) 프레임워크에서 심각한 보안 취약점 발견! CVE-2024-50050으로 추적되는 이 취약점은 CVSS 6.3점(9.3점으로 평가한 기관도 있음) 공격자가 악성 데이터를 역직렬화하여 임의 코드 실행(RCE)을 유발할 수 있음. 👉 영향: 라마 스택 추론 서버에서 원격 코드 실행 가능 https://t.co/TjfoCibHQv https://t.co/aQWpsrcIK8

    @zk_soulyong

    26 Jan 2025

    52 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    1 Reply

    0 Quotes

  42. Le Framework Llama de META à une faille de sécurité d'exécution de code à distance. La faille à la référence suivante : CVE-2024-50050. Plus de détail dans le lien si dessous 👇 https://t.co/dAmZ6MbSvp

    @_Nidouille_

    26 Jan 2025

    2265 Impressions

    15 Retweets

    28 Likes

    9 Bookmarks

    1 Reply

    0 Quotes

  43. CVE-2024-50050: Critical Vulnerability in meta-llama/#llama-stack by Meta - a popular #GenAI framework. The deserialization flaw allows remote attackers to execute arbitrary code, posing severe risks to AI model hosting, data integrity & system security: https://t.co/bY15s2h

    @securestep9

    26 Jan 2025

    4204 Impressions

    15 Retweets

    37 Likes

    7 Bookmarks

    1 Reply

    2 Quotes

  44. CVE-2024-50050: Critical Security Flaw in Meta’s Llama-Stack Framework https://t.co/Mdzk9YmxxD

    @Dinosn

    26 Jan 2025

    3079 Impressions

    6 Retweets

    26 Likes

    4 Bookmarks

    1 Reply

    0 Quotes

  45. Top 5 Trending CVEs: 1 - CVE-2024-49138 2 - CVE-2024-43468 3 - CVE-2024-50050 4 - CVE-2025-20156 5 - CVE-2020-11023 #cve #cvetrends #cveshield #cybersecurity https://t.co/4Fua3CAN6W

    @CVEShield

    26 Jan 2025

    167 Impressions

    1 Retweet

    3 Likes

    1 Bookmark

    0 Replies

    0 Quotes

  46. CVE-2024-50050: Critical Security Flaw in Meta's Llama-Stack Framework Stay informed about the critical vulnerability in Meta's Llama-Stack. Learn how this flaw can lead to arbitrary code execution and severe risks to AI model hosting and system security. https://t.co/YQM3gTVhQE

    @the_yellow_fall

    26 Jan 2025

    1607 Impressions

    8 Retweets

    29 Likes

    8 Bookmarks

    0 Replies

    1 Quote

  47. #exploit 1. CVE-2024-43451: Windows NTLMv2 0-day https://t.co/TgQztpU3HB 2. CVE-2024-50050: Vulnerability in meta-llama/llama-stack https://t.co/X1INA4cuA3

    @ksg93rd

    24 Jan 2025

    213 Impressions

    0 Retweets

    5 Likes

    2 Bookmarks

    0 Replies

    0 Quotes

  48. #exploit 1. CVE-2024-43451: Windows NTLMv2 0-day https://t.co/sTsJBKmCqh 2. CVE-2024-50050: Vulnerability in meta-llama/llama-stack https://t.co/F3EHLdhuUv

    @akaclandestine

    24 Jan 2025

    1570 Impressions

    14 Retweets

    34 Likes

    12 Bookmarks

    0 Replies

    0 Quotes

  49. 🚨 Critical Vuln Discovered in Meta's Llama Stack (#CVE-2024-50050) Attackers can now execute arbitrary code on inference servers via unsafe data deserialization. Full details: https://t.co/zXtvzhm1iw #Cybersecurity #GenAI #InfoSec

    @OligoSecurity

    23 Jan 2025

    252 Impressions

    1 Retweet

    5 Likes

    0 Bookmarks

    0 Replies

    1 Quote

  50. CVE-2024-50050 Llama Stack prior to revision 7a8aa775e5a267cf8660d83140011a0b7f91e005 used pickle as a serialization format for socket communication, potentially allowing for remote… https://t.co/PlnSaUZ20c

    @CVEnew

    23 Oct 2024

    328 Impressions

    0 Retweets

    0 Likes

    0 Bookmarks

    0 Replies

    0 Quotes

References

Sources include official advisories and independent security research.