Privacy Preserving IoT
summary
Argues that IoT privacy is recoverable with concrete tools and prototypes: uses Google’s Private Set Intersection (private join and compute) for matching without revealing raw records, Microsoft’s Homomorphic Encryption (SEAL) to run sums/means and even neural nets on encrypted inputs, and personal vaults like Hub of All Things (HAT)/DataSwyft plus the Sam project. Describes a Sam deployment that queries ~150,000 records across ~600 silos, runs computation enclaves, pays data owners, and is building a mobile app — all working end to end.
AI-written summary · about
chapters (7)
- 0:00Start
- 3:14Three Laws of Iot
- 14:57What Data We Share or Collect
- 21:38Symmetrical Encryption
- 23:15Alternatives to Centralized Data Storage
- 26:38Ethical Tech
- 42:40Neural Net Processing on Encrypted Data
related talks
39mData Privacy Engineering to power Responsible Enterprise AI
29mPrivacy-preserving methods: Building secure projects
26mPrivacy in the Age of IoT
50mSecuring IoT Data with Pervasive Encryption
25mImproving Safety with Privacy: An Exploration of Privacy...
29mIntro to Privacy-Enhancing Technologies (PETs)
01h 19mWhen Data Can’t Move but Insight Must
27mMetadata-private data transfer