Translate dialog fixes

[juliusknorr]

For the EU AI act we need to implement transparency requirements at each place where a user interacts with an AI system and in each AI generated content (like AI generated text, pictures, or audio). In each UI where the software interacts with an AI, display a warning like “This output was generated by AI. Make sure to double-check and adjust.”

The current dialog also has a few UI issues and could need some polishing

However given that for all other parts we reuse the assistant UI, I would suggest to check with the integrations team to see if we can also just use their UI (assuming it is somewhat exposed as a javascript API), if that is the case we should switch to it, if not we should add the note on AI output and polish our dialog.