The issue is not whether Europe should support AI innovation; it is whether AI systems that demonstrably fail safety standards and pose a risk to society should enjoy access to the European market