The scandal is that propublica wrote a story which directly contradicts their own statistical analysis. Their analysis was unable to show statistically significant bias (and had other flaws that might be genuine mistakes - no multiple comparison correction, possible misinterpretation of the model).

https://www.chrisstucchio.com/blog/2016/propublica_is_lying....

In fact, their r-script suggests that the secret algorithm is probably reducing racism relative to whatever secret algorithm human judges would use.