After the 2026 amendments, the wording focuses on supporting AI literacy rather than certifying one universal level for everyone. Practical needs depend on role, system and risk.
What should an AI user know, and what more should someone making decisions through AI understand?