My opinion is that they should not be if they were created using public data that is itself not owned by the creator/trainer. Otherwise we have a new model of f

My opinion is that they should not be if they were created using public data that is itself not owned by the creator/trainer. Otherwise we have a new model of for profit piracy where huge companies exploit their capital advantage to dragnet the collective work of humanity and then claim ownership of the result. If a model is trained on data that one owns the copyright to, it should be copyrightable.