I often use Tribler for torrents. It’s a TOR-like system specialized into torrent, and does work well with any torrents. (I’ll put a warning that the system might not be totally safe against targetted attack, but it should be against standard complaint to ISP)
If said content contain books (or maybe others pdf/epubs/docx), I would recommend uploading to one of the website that will eventually be mirrored by Anna’s Archive. (see https://annas-archive.org/datasets)
ISP can’t block it easily if it’s on I2P, something akin to TOR but also kinda different. Thought going against the hoster is totally possible if not anonymous (and there actually are already piracy (torrent over I2P) website on that network)
That’s actually a pretty good idea, thought It have some defect, in particular, each ActivityPub server have a limited view of the whole network. While it is usefull for avoiding abuse, it also have the downside that you can’t search for the whole thing that’s published on a platform.
But that could be solved with what is called backfilling (that Matrix does incredibly well). Sepia search (for Peertube) also does this.
Mixing ActivityPub with backfilling would be a really good idea. You can share metadata of ressources, have multiple instance, admin could block abusive website, and searching the whole site would be possible.
Maybe I’ll go study what already exist on this side.
(as an aside, Tribler does something similar to that, but only for Torrent and P2P)
I can confirm I have already some experience with the fact of VPN usage being flagged as a high change of automated traffic (except it was TOR, which is pretty much identical in this context).
Discord put me a wealth of captcha, Wikipedia refused edition of pages (even with my account. Which IMO looks like an oversight). And many pages just had captchas even when not trying to log-in.