Comment by kodah

The problem seems to be that spam (and fraud) are increasing, especially in the domain of identity.

Companies have been answering this growth with machine learning and that machine learning appears to scale poorly. Humans also scale pretty poorly. What would regulation look like?