Top Tweets for #LLMJacking
Critical flaws in 282 #iOS AI apps expose API keys and auth tokens, enabling #LLMjacking: unauthorized LLM requests billed to developers' accounts. Unauthenticated backend access also revealed proprietary system prompts, raising IP theft concerns.
LLMjacking just evolved. Sysdig caught a threat actor using an exposed Ollama server as the brain for an autonomous VAPT pipeline that fingerprints services and crafts exploits, all on stolen AI compute.
🔗 https://t.co/WJR2GyXiyW #Aisecurity #LLMjacking #CloudSecurity #LLM
113K+ requests in one month.
That’s how aggressively attackers targeted a fake AI server in a recent #LLMjacking experiment.
Threat actors are already scanning exposed AI servers, agents, MCPs, and RAG systems to hijack compute power and abuse credentials.
AI infrastructure security is becoming essential.
More:
https://t.co/xunxVb4l0Y
📢⚠️🤖 Operation Bizarre Bazaar has logged 35,000 attacks on exposed AI systems, stealing compute power and reselling access via the underground platform Silver Inc.
Read: https://t.co/ACcKlzwEsY
#CyberSecurity #AI #LLM #BizarreBazaar #LLMjacking
We turned our threat research into a game! 🎮
#GhostShell drops you into an AI pipeline under attack, inspired by real #LLMjacking techniques uncovered by Sysdig's Threat Research Team. Will you make the right choices & slip out undetected? Play it here ⤵️https://t.co/HY2lEMKMl0
サイバー犯罪者が“WormGPT”“FraudGPT”などジャイルブレイク済LLMを月$70‑$127で購入し、フィッシング文・マルウェアコード・攻撃手順を自動生成。AIにより脅威の民主化進行中。#WormGPT #LLMjacking
https://t.co/276Fnsjxr4
What is "LLMjacking" in cybersecurity? 🤔
A) Hijacking a live machine learning model
B) Attacking location-based services
C) Exploiting large language models
D) Encrypting network traffic unnoticeably
#Cybersecurity #LLMjacking #Infosec
. @Microsoft has named key developers behind malicious tools used to bypass AI safeguards in @Azure OpenAI Service.
🔗 Read more: https://t.co/r9FZqWC2kJ
Thanks to the experts from @oasissec and @keepersecurity for their expert insights
#AIAbuse #LLMJacking #ISBNews
🚨 Microsoft Discupts Storm-2139 for LLMjacking & Azure AI Exploitation - Hackers hijacked AI models using stolen API keys.
🔗 https://t.co/4bsyJbC3ox
#CyberSecurity #AI #LLMjacking #Microsoft
LLMJacking nimmt verstärkt Deepseek ins Visier
#Cybersecurity #DeepSeek #GenAI #künstlicheIntelligenz #LLM #LLMjacking #Security #Sprachmodell @Sysdig
https://t.co/4zmnZ4yHHa
Cybercriminals are cashing in on stolen AI credentials, targeting models like DeepSeek in a wave of LLMjacking attacks. @sysdig TRT’s research reveals how attackers exploit OpenAI Reverse Proxy (ORP) servers.
🔗Read more: https://t.co/HjomOD1Auq
✍️Kirsten Doyle
#AI #LLMjacking
🚨 LLMjacking is on the rise! Attackers steal API keys to access LLMs like OpenAI & DeepSeek—sticking victims with massive bills.
💰 One AWS user’s $2/month bill hit $20K in hours!
#CyberSecurity #LLMjacking #AIThreats #DeepSeek
https://t.co/N7ba5I1RTF
Big thanks to Lalit Choda and the NHI Mgmt Group for referencing our research on how threat actors compromise non-human identities (#NHI) for #LLMjacking campaigns.
https://t.co/IpIEMKNjrL
Big thanks to @Frichette_n for including our threat research on #LLMjacking in @HackingthCloud's Year in Review.
https://t.co/6GpyVocrhK
Thanks to the team at @wiz_io for referencing our research on #LLMjacking in their latest blog post!
https://t.co/hNwK8qOrk1
LLM-Jacking-Angriffe: Alarmierende Zunahme Mit Zunahme von LLMs (Large Language Models) nimmt auch die Bedrohung durch LLM-Jacking-Angriffe zu. #available_in_English #KITechnologien #LLMJacking #LLMs #Sysdig #Zugangsdaten
https://t.co/hDagymKXry
🚨 Atomic Fridays: #LLMJacking Indicators🚨
Here are a list of atomic #indicators related to recent attacks against #GenAI infrastructure. Keep an eye on your #NHI's:
--IPs--
104.28.154[.]76
122.53.249[.]18
156.228.110[.]211
...and 15 more. Full list in comments
--User Agents (UAs)--
Boto3/1.29.7
Python/3.12 aiohttp/3.10.5
python-requests/2.31.0 auth-aws-sigv4/0.7
...and more.
--ASOrgs--
Philippine Long Distance Telephone Company
CLOUDFLARENET
Datacamp Limited
...and others.
*Bonus snippet of an attacker script with one of the hardcoded UA's that was being used!
![TekDefense's tweet photo. 🚨 Atomic Fridays: #LLMJacking Indicators🚨
Here are a list of atomic #indicators related to recent attacks against #GenAI infrastructure. Keep an eye on your #NHI's:
--IPs--
104.28.154[.]76
122.53.249[.]18
156.228.110[.]211
...and 15 more. Full list in comments
--User Agents (UAs)--
Boto3/1.29.7
Python/3.12 aiohttp/3.10.5
python-requests/2.31.0 auth-aws-sigv4/0.7
...and more.
--ASOrgs--
Philippine Long Distance Telephone Company
CLOUDFLARENET
Datacamp Limited
...and others.
*Bonus snippet of an attacker script with one of the hardcoded UA's that was being used!](https://pbs.twimg.com/media/GccSLqHX0AAW5Qh.jpg)
Beutezug durch die Cloud
#CloudSecurity #Cybersecurity #KIResourceJacking #Kryptomining #LargeLanguageModel #LLMJacking #OpenSource #Security @Sysdig #ThreatResearch
https://t.co/Ap3JbO3da2
Wachsende Bedrohung durch LLM-Jacking
#CloudSicherheit #Credentials #Cybersecurity #KIModell #KITechnologie #künstlicheIntelligenz #LargeLanguageModel #LeastPrivilege #LLM #LLMJacking @Sysdig #ThreatResearch #Zugangsdaten
https://t.co/DvwqzoFiZx
This is some crazy research by @permisosecurity's very own @TekDefense around the darker side of #LLMJacking that our team has observed & investigated over the past 6 months. Protect your identities, control access to unused services & monitor those cloud spend alerts/bills!
🕵️♂️ LLMJacking for illicit roleplay 🕵️♂️
- GenAI infrastructure has quickly risen as a target for attackers
- With full prompt logging enabled, Permiso got a peek into what the hijacked models are being used for
- Spoiler: All for roleplaying of an illicit nature
*link in thread
Last Seen Hashtags on Sotwe
NoLimit() filter:videos
Seen from Portugal
รับงานนวนคร
Seen from Thailand
hoe sisters
Seen from United States
xxxdrones
Seen from Thailand
bokephijab
Seen from Singapore
剑三海鳗
incestoconpapá
Seen from Mexico
xlii() or #exny() or #momson() or #nolimit() +filter:native_video
Seen from United States
nolimit() filter:videos
Seen from United States
รีทวีตส่งคลิปเต็ม
Seen from Thailand
Trends for you
Most Popular Users
Elon Musk 
@elonmusk
240.7M followers
Barack Obama
@barackobama
119.2M followers
Donald J. Trump
@realdonaldtrump
111.7M followers
Cristiano Ronaldo
@cristiano
110.7M followers
Narendra Modi
@narendramodi
107M followers
Rihanna
@rihanna
97.7M followers
NASA
@nasa
92.2M followers
Justin Bieber
@justinbieber
90.9M followers
KATY PERRY
@katyperry
87.7M followers
Taylor Swift
@taylorswift13
81.6M followers
Lady Gaga
@ladygaga
73.1M followers
Virat Kohli
@imvkohli
70M followers
Kim Kardashian
@kimkardashian
69.8M followers
YouTube
@youtube
68.7M followers
Bill Gates
@billgates
63.9M followers
Neymar Jr
@neymarjr
62.8M followers
The Ellen Show
@theellenshow
62.4M followers
CNN
@cnn
61.9M followers
X
@x
60.8M followers
Selena Gomez
@selenagomez
60.8M followers