Languages: English | Français
A capability with which an attacker can issue queries to a trained machine learning model and obtain predictions or generations.
Source: NIST AI 100-2e2025 | Category:
