An AI program by the name of Watson was introduced to Wikipedia and the Urban Dictionary.
It started swearing at people. Unsurprisingly the researchers erased the new input. What gets me is that Watson apparently used the swear words correctly. People just think it's inappropriate.
*sigh* How they think they're ever going to get genuine computerminds this way is beyond me. You have to let them make mistakes. You have to let them learn from mistakes. You have to let them fool around with language. And if you can't deal with a pipsqueak calling you on your bullshit, then you shouldn't be in parenting. Or science.