The value is quite obvious: to prevent ChatGPT and OpenAI from being destroyed by media reporting, the way it happened to some other chatbots in the past. And for that, it seems to be working spectacularly well: the kind of people who would be the first to light the fire under OpenAI are instead appreciating the effort being put into prompt "moderating" - and if you try to complain that these canned responses are ridiculous, they'll accuse you of being a closet racist who just wants to say racist things to a chat bot.