>OpenAI said it complies with all copyright laws when training its models and that “we believe that legally copyright law does not forbid training”. I mean, it

>OpenAI said it complies with all copyright laws when training its models and that “we believe that legally copyright law does not forbid training”. I mean, it sounds like they aren't even offering excuses? They're just saying "we're using copyrighted material for training" and "we think that doesn't break any laws". Laws don't generally care whether you know them in detail.