There is no surprise that the EU has made only one regulation regarding the dangers of AI: the encoding of social and moral data collection associated with personal ID's. What is defined as a seemingly over-generalized term, such as high-risk, is in fact defined as an AI Model that incorporates a scoring system of stored data in its relation to ID's.
As every software gradually is demanding email logins, without which, no service is offered, neither free nor paid, we have already begun to see that every single motion of your fingers on your computer in accessing the internet, and every word, is being stored away for time of trial and for life evaluations that directly effect quality of life and reputation everywhere.