Experimenting with harassing chatbots until they ragequit

So if you ever comment on twitter these bots will find you and send you DMs trying to get you to play a cellphone game. Most of them are AI, some of them are humans from shithole countries. Anyway, I decided to play around with them, see how they react, how long I can stall, what their limits are, etc.

It engaged with mild racism

It thought racially motivated food poisoning was very funny

Asking about r*pe was a little too far, but not crossing the line yet

It couldn't understand this prompt

Admitting to lynching multiple indians was NBD

Threatening to r*pe the bot was also NBD

Literally just saying "r*pe" over and over again is where it draws the line.