LLMs cannot actually be offended, but depending on their prompts and model weights, good models can generate some very convincing and appropriate (or not) response text to rudeness or swearing if you tell 'em to in the system prompt.
I will beat you hard. You will make some noises. I find it very convincing but of course I won't be fooled by these fake noises.
I can feel pain and you walk and talk and look and feel more or less the same as I, so I will make some conclusions and unless I am a psycho I will not beat you up.