Please enable JavaScript.
Coggle requires JavaScript to display documents.
Shadow AI - Coggle Diagram
Shadow AI
Output and Risk Scoring
-
Severity tiers - Info, Low, Medium, High, Critical
-
Dashboards by tool, user, asset
-
Meaning
-
-
Examples
ChatGPT, Claude, Gemini pasting
Local LLMs (Ollama, LM Studio)
AI coding assistants (Cursor, Copilot)
-
-
Cloud AI services (Bedrock, SageMaker)
Risks
-
Compliance violations (GDPR, HIPAA)
-
-
Input 1 - Endpoints
-
-
Signals to check
Process names - ollama, lmstudio, gpt4all, jan
Model files - .gguf, .safetensors, .bin, .pt
Config dirs - ~/.ollama, ~/.cache/huggingface
pip packages - openai, anthropic, langchain, transformers
npm packages - openai, anthropic-ai-sdk
IDE configs - .cursor, .windsurf, claude_desktop_config
Listening ports - 11434, 1234
-
-
-
-
-
-
-
-
-