The inputs to FICO are far from secret; just the specifics of mapping them to scores are. If you've got bureau data on a wide population, it's trivial to nearly

The inputs to FICO are far from secret; just the specifics of mapping them to scores are. If you've got bureau data on a wide population, it's trivial to nearly perfectly replicate them with standard machine learning models.