With a view to ensuring a level playing field and effective protection, in particular of fundamental rights, the competent market surveillance authority should take account of the cross-border dimension of AI systems and assess whether an AI system that is found to present a risk may have affected natural persons in several Member States.