Metadata-private data transfer
summary
Survey of practical methods for hiding who accesses what and how well they actually scale Tor and Nym style mixnets give strong unlinkability at the cost of latency while private information retrieval variants (replicated PIR, homomorphic, lattice, hybrid trees) and private set intersection let clients read or check keys without revealing which item is requested but trade computation and async latency. Multiparty tools (DKG, MPC, VDF) and FHE hardware roadmap improve availability and throughput. Working sets under ~1M items are feasible with existing building blocks, billions require custom designs.
AI-written summary · about
chapters (14)
- 0:00Metadata privacy motivation
- 1:53Metadata privacy challenges
- 3:37End-to-end encryption limits
- 6:27Mixnets for anonymity
- 8:53Private information retrieval basics
- 11:43Computational PIR with HE
- 13:30Lattice-based PIR advances
- 14:55Private set intersection at scale
- 17:22MPC and distributed key generation
- 19:32Time-lock encryption
- 20:58Apple Private Relay and MPC
- 21:42Scaling limits and blocks
- 22:46FHE progress and future
- 23:28Q&A
related talks
27mBuilding blocks for Network Privacy
20mDeep Dive into Privacy Enhancing Technologies
08mPrivacy and Security
46mPrivacy Preserving IoT
01h 19mWhen Data Can’t Move but Insight Must
58mPrivate information retrieval at 30: Achievements, disappointments, and open problems
31mTowards making private telemetry as ubiquitous as TLS
13mSurvey of Privacy Solutions for Minimizing MEV