Comment by jermaustin1

> [Unpaywalled] data should automatically have no copyright beyond authorship attribution.

Isn't that't the problem, no way to know exactly where it "learned" from? That is what they are fighting in the courts, I believe.