the only people responsible for this are whichever execs ordered the filtering/censoring. Getting a chat bot to say "naughty things" was doing no one any harm.
If I screen shotted me writing down a bunch of racists phrases via MS Word, would they patch the software to disallow that content?