I think you're reading it wrong. Regulations may have a provision that allows providers to apply age assurance [systems ?] if the age assurance is highly effective at determining age.
I'm always surprised how ambiguous the writing is for this kind of stuff. Maybe that's the point. If the regulations don't (may is optional) have the provision, does that mean they need to demand ID?
IMO, highly effective = our buddies' tech that we declare highly effective. The whole ID push around the world is big tech trying to set up government mandated services that you're going to be forced to pay for, either directly or via taxes.
The end game is probably digital IDs with digitally signed requests for everything you do. And, of course, corrupt individuals and criminals will somehow be able to get as many digital IDs as they want.
That money should be spent on education. We're being robbed.