Google General Purpose Transpiler for Fully Homomorphic Encryption

0



Google last month unveiled a one-of-a-kind general-purpose transpiler for fully homomorphic encryption (FHE). The new update allows developers to calculate encrypted data without having to access personally identifiable information. Open source libraries and tools for performing FHE operations on an encrypted dataset are available at GitHub.

In addition to this, there have been a lot of open source libraries and tools to perform FHE operations in the past, including TFHE, Concrete, HEAD, etc. Consult the Complete list resources, libraries and software for fully homomorphic encryption here.

Register now

“We are proud to be an industry leader in the development, deployment and scale-up of new privacy-preserving technologies that help create meaningful experiences and learn valuable information while by protecting the privacy of our users ” mentionned Miguel Guevara, product manager, privacy and data protection office at Google, in a blog post.

How does the transpileur work?

With FHE, encrypted data can travel over the Internet to a server, where it can be processed without being decrypted. The transpiler will allow developers to write code for any kind of basic calculation (be it simple string processing or math) and run it on encrypted data. This allows developers to create new programming applications that do not need unencrypted data. In addition, FHE can be used to train machine learning models on sensitive data in private.

The design of Google’s transpiler is modular in three ways: (as shown in the image below)

Client-server interaction (Source: arXiv)
  • The ‘entry code’ can be in any language which can be translated to XLS. This reduces the burden of transpiling existing code as it does not need to be written in a fixed supported language.
  • The ‘exit FHE code’ can be in any language with an FHE library. This reduces the need to interact with “transpiled FHE code” because “FHE code” can be transpiled into a “language” that interfaces well with the rest of the code.
  • The underlying “FHE backend” can be any library that exposes gates as part of its API. “Our library includes classes prefixed by Encode, which can be reused for ease of development. This can speed up FHE search by providing a simple way to compare arbitrary programs side-by-side in different FHE patterns, ”according to Google.

Take the example of creating an app for people with diabetes. This app can collect sensitive data from its users, and you need a way to keep this “private and protected data” while sharing it with medical experts to gain valuable information. With transpiler for FHE, you can encrypt the data you collect and share it with medical experts who, in turn, can analyze the data without decrypting it – providing useful information to the ‘medical community’, while ensuring that that no one can access the underlying personal information.

Cite a healthcare startup Arkhn, Google said the company could accelerate scientific discovery by using differential privacy to share data between hospitals.

Four years ago, Google researchers invented Federated learning, which helps maintain privacy by keeping as much personal information as possible on your device. Two years ago, Google made its privacy differential library available for free to any organization or developer. This is an anonymization technology that allows developers to learn about their data in private.

Currently, data anonymization is the most widely used technique, especially in the health sector. This is a process of removing all personally identifiable information from the dataset while retaining only the relevant portion without compromising user privacy. For example, hospitals and clinics typically remove names, addresses, phone numbers, and other vital patient information from health records before incorporating them into large data sets.

See also
Optimization

Theoretically, anonymizing data might seem like a great idea. But, in reality, it is not completely foolproof because anonymization does not resist de-anonymization attacks, since it is often linked to auxiliary information to identify the persons concerned.

What’s in store?

Over the next ten years, Google said that FHE could even help researchers find associations between specific genetic mutations by analyzing genetic information on thousands of “encrypted samples” and “testing different hypotheses” to identify them. genes associated with the diseases they are studying.

“We still have a ways to go before most of the computation takes place with FHE, but as long as it took a while for HTTPS to take off and be widely adopted,” according to Google.

Google has said the launch of its fully homomorphic encryption is a step towards delivering useful products that maintain user privacy and protect their data.


Join our Telegram group. Be part of an engaging online community. Join here.

Subscribe to our newsletter

Receive the latest updates and relevant offers by sharing your email.

Amit Raja Naik

Amit Raja Naik



Amit Raja Naik is Editor-in-Chief at Analytics India Magazine, where he dives deep into the latest technological innovations. He is also a professional bass player.



Leave A Reply

Your email address will not be published.