Zero-Knowledge Federated Learning with Lattice-Based Hybrid Encryption for Quantum-Resilient Medical AI
Edouard Lansiaux · 2026
Read Zero-Knowledge Federated Learning with Lattice-Based Hybrid Encryption for Quantum-Resilient Medical AI by Edouard Lansiaux (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-03398.txt.
Description
Edouard Lansiaux designs a safer way for hospitals to train medical artificial intelligence together. Federated learning lets each hospital keep its patient data while sharing model updates. But those updates leak: attackers can reconstruct patient information from them, dishonest clients can poison the shared model. Traffic encrypted today may also be read by future quantum computers. His 2026 protocol, ZKFL-PQ, layers three lattice-based defenses against all three threats.
Quantum-resistant key exchange protects the channel, zero-knowledge proofs verify that each update respects norm limits without revealing it, and homomorphic encryption lets updates be combined while still encrypted. Security proofs rest on standard lattice assumptions. On synthetic medical images with five clients, the protocol rejected every oversized malicious update and kept accuracy perfect, where plain federated learning collapsed. The heavy computational cost fits daily or weekly research cycles, the author argues.
Catalog record
| CREATOR | Edouard Lansiaux |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31241 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.03398