AI
Breaking Wire
Content
Ecosystem
Community
Pricing
AI Vulnerability Database
926 vulnerabilities tracked — CVEs, prompt injection, model exploits.
MEDIUM
Privilege Escalation in AWS SageMaker Studio due to Overly Permissive Default IAM Roles
AWS SageMaker Studio with default IAM role configurations
10 Sept 2025
0 views
Unpatched
HIGH
Cross-Domain Data Exfiltration in AI Assistants via Indirect Prompt Injection
Previous
Page 50 of 78
Next
AI Agents with Web Browsing Tools
Email Summarization Services
Document Analysis Chatbots
Multi-modal LLM applications
10 Sept 2025
0 views
Unpatched
HIGH
CVE-2024-0073
NVIDIA GPU Driver Kernel Mode Layer Allows for Privilege Escalation in Multi-Tenant AI Clusters
NVIDIA GPU Driver (Linux) < 551.81
NVIDIA GPU Driver (Windows) < 555.52
10 Sept 2025
3 views
Patched
HIGH
GitHub Copilot Suggests Vulnerable Code Snippets Leading to Common Security Flaws
GitHub Copilot
Amazon CodeWhisperer
Tabnine
10 Sept 2025
25 views
Unpatched
HIGH
Data Exfiltration via Indirect Prompt Injection in LLM-Powered Document Analysis Tools
Any LLM application processing external unstructured data
LangChain agents
AI-powered chatbots
RAG pipelines
10 Sept 2025
10 views
Unpatched
HIGH
CVE-2023-49080
Malicious PyPI Package 'torchtriton' Exfiltrates Sensitive Data from AI Developer Environments
Python developer environments
CI/CD pipelines using PyPI
10 Sept 2025
5 views
Patched
HIGH
Indirect Prompt Injection in Web-Browsing AI Agents Enables Session Takeover and Data Exfiltration
AutoGPT
LangChain agents with browsing tools
LlamaIndex query engines with web retrieval
Any LLM-powered agent with internet access
10 Sept 2025
5 views
Unpatched
MEDIUM
Data Extraction via Targeted Fine-tuning Attack on AWS Bedrock Titan Models
AWS Bedrock (Amazon Titan Text G1 - Lite)
AWS Bedrock (Amazon Titan Text G1 - Express)
10 Sept 2025
5 views
Unpatched
MEDIUM
Codebase Exfiltration via Malicious Workspace Content in GitHub Copilot
GitHub Copilot for Visual Studio Code (< 1.185.0)
10 Sept 2025
9 views
Patched
HIGH
Indirect Prompt Injection in AI Agents via Unsanitized Web Content
AI-powered web browsers
Autonomous AI agents
Email summarization tools
Document analysis chatbots
10 Sept 2025
18 views
Unpatched
HIGH
GitHub Copilot Context-Awareness Abused to Suggest Insecure Code Containing Sensitive Data
GitHub Copilot for VS Code < 1.185.0
JetBrains Copilot Plugin < 1.5.0
10 Sept 2025
48 views
Patched
CRITICAL
Malicious AI Models on Hugging Face Hub Execute Code via Unsafe Pickle Deserialization
huggingface-transformers
pytorch
any library loading models from .pkl or .bin files
10 Sept 2025
13 views
Patched