Data used to evaluate the performance of a final AI model (in general) or machine learning model (in particular) before it goes live.

Note: Basically, test data shall be disjoint from training data and validation data.

ISO/IEC DIS 22989 data used to assess the performance of a final machine learning model (3.2.11).

Note 1 to entry: Test data is disjoint from training data (3.2.22) and validation data (3.2.24).

ISO/IEC TR 29119-11:2020 independent dataset used to provide an unbiased evaluation of the final, tuned ML model (3.1.46)