The practical exam point: transparency duties exist, but users still need verification habits.
Article 50 requires clear labelling of AI-generated content and deepfakes; the exam trap is confusing this with broader transparency rules under Articles 13 and 52.