License files (and copyright) aren’t omnipotent. I understand why you might feel the ethics lie in a different direction, but the established legal precedent is

License files (and copyright) aren’t omnipotent. I understand why you might feel the ethics lie in a different direction, but the established legal precedent is that parsing and building search indexes or neural nets is generally fair game regardless of what the copyright holder wants. You mentioned robots.txt - there’s no legal enforcement behind that either. Notably, the Internet Archive bot ignores it completely.