>So when a plaintiff using a hiring platform encounters a problematic design feature — like platforms that check for gaps in employment — she should be able to

>So when a plaintiff using a hiring platform encounters a problematic design feature — like platforms that check for gaps in employment — she should be able to bring a lawsuit on the basis of discrimination per se, and the employer would then be required to provide statistical proof from internal and external audits to show that its hiring platform is not unlawfully discriminating against certain groups. How would this work? Without ground truth labels of "should have been hired" and "shouldn't have been hired" how can you demonstrate that your algorithm isn't biased? I mean counterfactual labels like "we hired this person the algorithm said not to hire, and they turned out good/bad." Base rates will be different, which means any algo will be biased in some way (see these impossibility theorems [0]), so the question is what demonstration of non bias will be sufficient without counterfactual data. [0] https://www.chrisstucchio.com/pubs/slides/crunchconf_2018/slides.pdf https://www.chrisstucchio.com/pubs/slides/crunchconf_2018/sl...