causal
Fairness Approaches Ignore Historical Root Causes
Much algorithmic fairness/accountability/transparency work attempts to tune bias out of a closed system starting from time-zero, which fails because it omits root-cause analysis of how, for example, lending to Black Americans came to be perceived as high-risk through centuries of deliberate discrimination (redlining, racial covenants); a genuine equity-and-justice approach must build the past 200 years of history into the model rather than treating it race- and gender-blind.
causalpending
Speaker
Catherine D'IgnazioEvidence Quote
“We have to acknowledge that past, and we have to account for it somehow in our systems”
Created: 6/13/2026, 12:14:23 AM
Your notes
Loading notes...