The reporting here implies that devices for customers who don't accept the new terms may "cease to function". Does the mechanism for how that happens matter? The bottom line is that either the product continues to provide whatever functionality it was purchased for, or it's now broken.

Typically under consumer protection laws in places like Europe, a product that ceases to work properly unreasonably quickly after it is purchased will trigger certain rights for the purchaser, such as getting the device repaired or replaced so it works again, or getting a full or partial refund as compensation.

Importantly, there is typically no exception to this principle just because the product in question involves software or depends on remote services. In fact, the EU strengthened consumer protections on digital purchases not so long ago, so the direction of the legal currents is very clear.

I suppose you could say there has been some sort of informal understanding that because it's so hard to write software with no bugs or security flaws, it's reasonable for the suppliers of products with software components to supply software updates post-sale to keep things working properly without falling foul of the basic "it no longer works" trap if users don't then apply those updates. But the sorts of shenanigans we've seen recently with software developers bundling unrelated and possibly unwanted changes in with their bug fixes and security patches are stretching good faith here, and at some point I suspect something has got to give.

The same principle applies to products with dependencies on remote services that don't necessarily last as long as the product itself, whether that's a server for a multiplayer game, or a DRM scheme with remote authorisation, or a provider of audio data that makes a speaker system useful.