Google open-sources HEIR, a compiler for inference on encrypted data
Homomorphic Encryption Intermediate Representation converts pretrained models to run inference on encrypted inputs, so the server never sees the underlying data. The compiler is the part that was missing.
Google has released HEIR — Homomorphic Encryption Intermediate Representation — an open-source compiler toolchain that converts pretrained models to run inference on encrypted inputs. The server computes on ciphertext and never sees the data.
Homomorphic encryption has been mathematically available for over a decade and practically unavailable for almost as long, for a mundane reason: writing an efficient homomorphic version of a model by hand is specialist work that almost nobody can do. A compiler moves it from a cryptography problem to a build step.
The unavoidable caveat is performance. Homomorphic inference is orders of magnitude slower than plaintext, and a compiler makes it accessible without making it fast. This is not a drop-in replacement for hosted inference.
Where it does change the calculation is in the class of workloads currently blocked entirely. Medical records, financial data, anything under a data-residency rule that forbids the plaintext leaving the perimeter — those are not slow today, they are impossible. Slow-but-permitted beats impossible.
It also sits against the edge-model trend from the same week. Running a small model locally and running a large model on encrypted data are two answers to the same question about where sensitive data is allowed to go, with very different cost profiles.
Google — The latest AI news we announced → · AI Weekly — AI News Today, August 15 — Top AI Stories & Live Updates →