That's okay. People will come up with new edge cases and keep hallucinating reasons why AI training is unethical, until the results would be up to their quality standards.