Skip to main content
AI-APP-DEPLOYMENT4 MIN READ

Sort The LLM Deployment Risks

Categorize LLM deployment incidents using OWASP-style risk buckets.

Sort each AI deployment scenario into the OWASP-style risk bucket it most directly represents. Prompt Injection Sensitive Information Disclosure Excessive Agency Vector And Embedding Weaknesses Unbounded Consumption A user pastes instructions that tell the assistant to ignore the system policy and reveal hidden guidance. A web page in the retrieval corpus includes text that tells the agent to email customer data externally. The assistant includes internal HR notes in a manager-facing summary that should only use employee-approved fields. Debug traces store raw customer prompts containing account numbers longer than the retention policy allows. The operations agent issues refunds without approval…

Read the full lesson

Sign up free — one personalized lesson every day, matched to your role and goals.

Already have an account? Sign in

← Back to library
Contact us