Pierre Rochard talks about the four layers of defense in his recent conversation with Preston.

The first is the social layer - convince inscriptors to stop either through good faith debate (unlikely to work because they're probably bad faith actors) or ridicule (eg. #gorillawarfare)

The second is economic - the fee market will take care of it. If people value "real" bitcoin transactions more than spam, which is likely to happen as it becomes a more widely accepted form of money, then inscriptions will disappear naturally by being priced out

The third and fourth are nuclear options, IMO. Node level rejection or protocol/consensus level changes. Unlikely that it will come to this as I expect layers 1 and 2 to work. These options also fall into the grey area of censorship, so I'm not sure I'm 100% comfortable with them.