1 paper
Ho Wa Ku, Jameel Ahmed Siddiqui
AI governance and assurance often assume that a consequential model-mediated decision can be reconstructed or tested after the fact. That assumption may fail when the evaluator tha…