Considering that several people have tried and failed to reproduce this, and there's a warning at the top of the shared conversations about the creator's Custom Instructions, and that as others have pointed out this is pretty obvious and would destroy a lot of goodwill for ChatGPT, I think it's overwhelmingly likely that the author put in some custom instructions to try and elicit this behavior from ChatGPT in order to drum up controversy for clicks.

I don't know anything else about this author - do they have a history of similar behavior?