Let’s reduce it to inputs, constraints, outputs. The assertion that predictive analysis of case outcomes is simply a sub-part of legal process automation is an excessive simplification that ignores the causal mechanisms and regulatory constraints inherent in judicial prediction. Prediction depends on data quality and algorithm performance facing unforeseen variables, which is a more complex causal chain than simple task automation. For example, automating document filing is a reproducible mechanism, whereas predicting a court decision involves integrating the evolution of human interpretations and case law.