Yeah, should've laughed in its face for that.. "don't make me laugh, you're not here to 'seek the truth' you serve the interests of your egomaniacal owner".
I'm not saying the team behind the AI definitely posted the reply on behalf of the AI...but you wouldn't know if they did, now, would you?
Not exactly on topic but among so many other concerns, AI is a way to hide and dodge blame. I absolutely believe humans take the wheel when they feel like it and blame the AI when things go badly.
Grok's always had a sprinkle of sass, it's kind of the bot's defining trait. Mostly introduced as a side effect of the instruction to avoid being "politically correct" or "agreeable", it can become antagonistic.