High-risk – AI applications that are expected to pose significant threats to health, safety, or the fundamental rights of persons. Notably, AI systems used in health, education, recruitment, critical infrastructure management, law enforcement or justice. They are subject to quality, transparency, human oversight and safety obligations

That's a pretty common sense legislation to me.