I worry what this means for the future of deep-learning based audio analysis tools to automatically generate subtitles (for the hard of hearing). Does this mea

I worry what this means for the future of deep-learning based audio analysis tools to automatically generate subtitles (for the hard of hearing). Does this mean that YouTube's audio-recognition powered automatic subtitles are illegal in some edge cases? The platform can always have a ToS/shrinkwrap allowing for automated subtitling, but legacy videos created before the policy or videos that aren't owned by the uploader could cause legal problems. Also, is this possibly case law that can be used to outlaw other forms of describing media, such as drawing representations of the scenes or writing descriptions of what happens? Yes, that's pulling the current situation to obserb conclusions but sometimes that's what happens with copyright.... always trying to push the law in ways it was not intended.