< Schizo notes // FHE  >

This is going to be a long post, but if you fancy exploring the upcoming narrative/tech, it will be well worth it ⚠️

What if you don't need to decrypt data for computing tasks? FHE - the “Holy Grail of Cryptography”

Growth Lead @ RedStone | Degen | Micro-angel

1/ Any computational operation on data typically requires decryption, exposing sensitive user information to service providers. To enhance privacy and security, the adoption of end-to-end encryption is essential.

2/ Fully Homomorphic Encryption (FHE) offers a transformative approach to data security by allowing computations on encrypted data without the need for decryption.

3/ This technology not only preserves the confidentiality of data but also broadens the scope for secure data analysis and the development of decentralized services.

4/ By maintaining data encryption throughout processing, FHE significantly enhances privacy protections and opens new possibilities for secure and private computation in various fields.

5/ FHE is not yet widely adopted mainly due to its significant computational costs. This form of encryption generates extremely long bit strings, often exceeding 10,000 bits, which are not efficiently processed by standard GPUs and CPUs.

7/ Excitingly, companies such as intel, DARPA, Chain Reaction are in the final stages of developing specialized chips for FHE. These chips are tailored to address the computational challenges associated with FHE.

Personally, though, I see them as chips for AI or NPU -> the song of the future.

8/ Prime example of FHE application lies in its integration into AI systems. Consider this scenario: Before transmitting facial image data to the cloud, the client encrypts it. Then, the AI model at company X, equipped with FHE capabilities, processes these encrypted data.

9/ As a result, it generates encrypted results that only the client can decrypt using their private key. This approach ensures that sensitive data remains undisclosed and secure during processing in the cloud.

10/ Another prevalent application of FHE lies in the healthcare sector. Imagine securely transmitting your health data without the risk of it being compromised. This technology offers a fundamental solution to the longstanding privacy concerns surrounding sensitive medical information.

11/ FHE isn't merely a luxury or an optional feature; it's becoming an indispensable tool for safeguarding our most intimate data.

Currently, the utilization of FHE comes at a considerable cost, deterring many from its widespread adoption until optimization occurs.

12/ However, I believe there's one particular domain where users are not only willing but eager to pay substantial fees—an ideal testing ground. Does that ring a bell? 

Yes, I'm talking about you degen 🫵

13/ FHE x Web3 

Blockchain tech encounters a significant hurdle in maintaining data security during transactions. Fully Homomorphic Encryption offers a direct solution to this issue by enabling the execution of computations on encrypted data, eliminating the necessity to decrypt it first.

14/ Let's look at some of the uses of FHE in web3:

Secure Transactions: FHE enables users to encrypt transactions locally, enhancing security and privacy in blockchain interactions.

Encrypted Mempool: Transactions in an FHE-encrypted mempool are protected from unauthorized access, addressing issues like frontrunning.

15/ Encrypted Blocks: With FHE, entire blockchain blocks can be encrypted, safeguarding the confidentiality of the data they contain.

State Transition with FHE EVM: The FHE EVM allows for secure state transitions in blockchain systems without exposing sensitive data.

16/ Privacy by Design: FHE ensures transactions remain encrypted from wallet to blockchain, bolstering user data protection against unauthorized access.

17/ Currently, FHE is viewed not as a standalone solution but as a crucial piece in a privacy-preserving toolkit, expected to be integrated with technologies like ZKP, MPC, and TEE for enhanced security.

There are already projects that combine FHE and ZK, such as Pado Labs

18/ The smoothest synthesis of web3 and FHE is fhe EVM.

fhEVM, developed by Zama, is not a blockchain but an encryption tool for EVM-compatible blockchains, enhancing privacy through FHE.

19/ It includes a threshold protocol for FHE, a Solidity library for writing smart contracts on encrypted data, and a dApp SDK for creating applications that interact with encrypted data.

20/ How fhEVM Works

fhEVM encrypts all data using a global FHE public key, with the secret key distributed among validators via a threshold protocol. Validators perform computations on encrypted data and can decrypt or re-encrypt values using this protocol.

21/ This setup allows for confidential smart contracts and transactions while maintaining data availability and composability on-chain.

22/ Projects Integrating FHE and web3:

Zama: This open-source cryptography company is at the forefront, developing FHE solutions tailored for blockchain applications and ai.

Octra: A blockchain network specializing in FHE, offering support for isolated execution environments to safeguard data integrity and privacy.

23/ Fhenix: Pioneering the space as the first confidential Layer 2 solution that FHE to enhance security.

Shibarium:  A blockchain dedicated to memecoins that has successfully integrated Zama's advanced cryptographic technology.

24/ Privasea: Marking its place as the world's initial network to implement FHE for machine learning inference.

Inco: Positioned as the universal confidentiality layer for the web3 ecosystem, this platform combines the robustness of FHE with the security of Ethereum.

25/ Mind Network: Zero Trust Layer of Web3 featuring an FHE Restaking Layer for AI and DePin networks

26/ FHE is a real thing, and projects seeking to implement them are currently attracting record levels of funding. Zama recently raised over $73 million from funds like Multicoin Capital!

26/ FHE is a real thing, and projects seeking to implement them are currently attracting record levels of funding. Zama recently raised over $73 million from funds like Multicoin Capital! 

https://x.com/zama_fhe/status/1765727386480320771

27/ Final Words:

Personally, I think FHE is an important piece of the puzzle for the advancement of privacy technology. I also think it is one of the most important pieces for AI, and it dovetails perfectly with the use of blockchain.

28/ I also agree with Zama's CEO that web3 will be a testing ground for the use of FHE in practice. I think the upcoming zkFHE type solutions will be complementary encryption methods.

29/ Want to go deeper into the FHE rabbit hole?

I suggest beginning with Taiko insightful article. It serves as an excellent foundational resource to enhance your understanding of FHE:

29/ Want to go deeper into the FHE rabbit hole?

I suggest beginning with Taiko insightful article. It serves as an excellent foundational resource to enhance your understanding of FHE: https://taiko.mirror.xyz/2O9rJeB-1PalQe

6/ As a result, operations involving FHE tend to be slow and energy-intensive. However, despite these challenges, FHE holds potential for future-proofing data security against quantum computing threats, as its complex algorithms could be resistant to quantum attacks.