US courts are increasingly using algorithm-based recidivism risk prediction instruments in estimating offenders’ dangerousness and, thus, the warranted severity of the punishment. Some argue that this practice mitigates well-known biases in non-algorithm-based recidivism risk assessments. Whether this is so or not, in the present US context, algorithm-based sentencing is quite likely to be unfairly discriminatory. This claim might have radical implications regarding the US penal system in general (as well as that of all other countries); to wit, that it too is unfairly discriminatory against groups with high base crime rates. If so, we face a dilemma forcing us to revise some of our beliefs about penal justice.