This series reviews prior art for Match Feed, a design for federated video recommendations in which people choose specific peers and exchange encrypted taste data, with ranking done on their own devices. It covers decentralized and consented recommender systems, how new creators get discovered, how states have pressured federated networks and what relays and P2P swarms reveal about viewers, plus a log checking each source the design cites. The literature review closes with a simulation update showing which of its recommendations held up.
The pieces refer to "DESIGN.md": that is the Match Feed design document in the project's public repository. All of the research was done and checked on October 19, 2026.
The literature review is filed on the AI & society shelf and the censorship and metadata report on the Cybersecurity & digital safety shelf. They are written to be read together.