Technical Evaluation of a Disruptive Approach in Homomorphic AI

TL;DR

This paper evaluates HbHAI, a novel cryptographic approach enabling AI algorithms to operate directly on secure data with high performance, confirming its security and efficiency through various tests.

Contribution

It introduces HbHAI, a new hash-based homomorphic method that allows native AI algorithms to process encrypted data without modifications, offering improved performance over existing schemes.

Findings

HbHAI preserves similarity properties suitable for AI

AI algorithms perform effectively on HbHAI-protected data

Security and operability are largely confirmed with minor reservations

Abstract

We present a technical evaluation of a new, disruptive cryptographic approach to data security, known as HbHAI (Hash-based Homomorphic Artificial Intelligence). HbHAI is based on a novel class of key-dependent hash functions that naturally preserve most similarity properties, most AI algorithms rely on. As a main claim, HbHAI makes now possible to analyze and process data in its cryptographically secure form while using existing native AI algorithms without modification, with unprecedented performances compared to existing homomorphic encryption schemes. We tested various HbHAI-protected datasets (non public preview) using traditional unsupervised and supervised learning techniques (clustering, classification, deep neural networks) with classical unmodified AI algorithms. This paper presents technical results from an independent analysis conducted with those different, off-the-shelf AI algorithms. The aim was to assess the security, operability and performance claims regarding HbHAI techniques. As a results, our results confirm most these claims, with only a few minor reservations.