Yes, but animations have to be loaded by the client in the first place. Models are different because they are instance content that can only get loaded by the server, animations have to be able to be loaded by the client.
This is just making an API call, but in the browser. When I say API call, I mean via HTTP requests to the different asset delivery endpoints that Roblox uses to fetch asset content. You can’t make something private without actually making it private, you can’t have animations not be loadable by the client without making animations not loadable by the client.
If you insert a model into a game it can be stolen in the exact same way that an animation can, the difference is that an animation has to be loaded by all clients so that they know the keyframes to play back in the first place.
You can’t get rid of this without getting rid of the client’s ability to play the animations unless you are suggesting that the server should be downloading the animation content and then having connected clients then download it from the server, which has its own large set of flaws and complications for a fairly insignificant gain.
Notably, this would mean that all of this content delivery has to get routed through Roblox game servers which would be pretty inefficient and would ultimately slow down asset loading and, if you actually want to protect the assets, prevent you from being able to cache it at all.
Caching is also another way content can easily be stolen, I have even written tools experimentally that can rip unions and meshes out of the cache, and even recover the contents of scripts included in the parts that made up the unions since that data is part of the union (for the purpose of separating the union).
I am not really sure I understand what you mean by this at all, Roblox already does automatically check the contents of models.
Okay but that’s still definitely not a reason to justify sequential asset IDs, the ‘feeling’ that sequential asset IDs are somehow necessary here is a symptom to the greater problem of moderation tooling.