01/09Security

AI Security & Safety.

Prompt injection, OWASP LLM Top 10, output validation, data leakage prevention, and AI governance.

AI Security & Safety · 1 of 3
Prompt Injection
What classic software vulnerability is prompt injection most often compared to?
AI Security & Safety · 2 of 3
Responsible AI & Governance
The EU AI Act uses four risk tiers. Which tier requires transparency obligations such as informing users they are interacting with AI?
AI Security & Safety · 3 of 3
Data Leakage & Privacy
Someone with only API access repeatedly queries your facial recognition endpoint and uses the returned outputs to reconstruct training images of individuals. What technique is this?

Three of the 947 AI Security & Safety questions in the bank.