Being mean to a chatbot is like playing No Russian from Call of Duty: MW2. Of course it's horrific to do in real life, but it's not real life. We can use it to think about and reflect on real world issues, but the game itself is fine.

I can see two things that being mean to chatbots might cause ethical issues. The first is training real people in how they act in relationships. It's been shown that playing video games doesn't morally affect you much in real life, but this is interacting in the same way that we interact with real people. We know that people are meaner over the Internet than in real life. I'd like to measure this in some way before taking a stance on it.

The second is that chat bots are built using existing real-world conversations. Being mean to chatbots means the next generation of chatbots is mean, too. The last time one got exposed to the Internet, we had it praising Hitler within hours. It's not good for the chatbot industry, and ironically sabotaging work others produce sincerely might be considered ethically wrong.