Privacy is a fundamental right for both individuals and organizations. It allows people to freely express themselves without revealing information they do not wish to share to third parties. For many organizations today, data is considered a primary commodity, making data privacy crucial for protecting this commodity. The cypherpunk movement and data commercialization have accelerated research and development in cryptographic primitives.
Cryptology is a rather extensive field, and when viewed in the context of computing, we have seen various schemes emerge since the 1960s, such as zero-knowledge proofs, homomorphic encryption, secret sharing, etc., all of which have been continuously improved. These schemes are crucial for unlocking private computational methods (data is a primary commodity because insights can be derived from it). To this day, the field of Private Computing has made significant advancements in multi-party computation and zero-knowledge proofs, but privacy issues persist with input data itself.
When the most valuable commodity is disclosed, it is very difficult for any data owner to outsource the computation of this data without a legal agreement. Today, everyone relies on compliance standards for data privacy, such as HIPAA for health data and GDPR for data privacy in the European region.
In the blockchain field, we place more trust in the integrity of technology rather than regulatory bodies. As believers in permissionless and maximal ownership, if we believe that users own the future of data, we need trustless methods to compute on this data. Before Craig Gentry’s work in 2009, the concept of computing on encrypted data had not been achieved. It was the first time someone could perform computations (addition and multiplication) on ciphertexts (encrypted data).
Homomorphic Encryption (FHE) allows computations to be performed on encrypted data (ciphertexts) without decrypting the data, opening up a range of use cases for privacy and data protection. In the FHE process, when data is encrypted, additional data called noise is added to the original data. This is the process of encrypting data.
Each time a homomorphic computation (addition or multiplication) is performed, additional noise is added. If the computation is too complex and noise is added each time, decrypting the ciphertext will become very difficult (this is computationally intensive). This process is more suitable for addition, as noise grows linearly, while for multiplication, noise grows exponentially. Therefore, if there are complex polynomial multiplications, decrypting the output will be very difficult.
If noise is the primary issue and its growth makes FHE difficult to use, it must be controlled. This gave rise to a new process called “Bootstrapping.” Bootstrapping is a process of encrypting encrypted data using new keys and decrypting within the encryption. This is crucial as it significantly reduces the computational overhead and decryption costs. Although Bootstrapping reduces the final decryption costs, there is a significant amount of operational overhead during the process. This can be both expensive and time-consuming.
The main FHE schemes currently are BFV, BGV, CKKS, FHEW, and TFHE. Except for TFHE, the abbreviations of these schemes are the names of their paper authors.
FHE is more focused on outsourcing general computation in the blockchain field rather than building integrated FHE L1/L2 solutions. Some interesting use cases that FHE can unlock include first-generation (encryption-native): on-chain DIDs, casinos, betting, voting, games, private DeFi, private tokens, dark pools, 2FA, backup, and passwords; second-generation (modular): “privacy chains” (Chainlink for privacy), outsourced private computation, end-to-end encryption between blockchains and contracts, encrypted data availability, and verifiable secure data storage; third-generation (enterprise): complex consumer applications, encrypted and decentralized LLM, artificial intelligence, wearable devices, communications, military, medical, privacy-preserving payment solutions, and private P2P payments.
Projects in the industry based on FHE have sparked innovation in enhancing data privacy and security using this technology. This section delves into the technical details and unique approaches of notable projects such as Inco, Fhenix, and Zama.Lattice-Based FHE: Inco utilizes lattice-based encryption for its FHE implementation, known for its post-quantum security features, ensuring resilience against potential future quantum attacks.
Privacy-Preserving Smart Contracts: Inco’s smart contracts can execute arbitrary functions on encrypted inputs, ensuring that both the contract and the nodes executing the contract cannot access plaintext data.
Noise Management and Bootstrapping: To address the issue of noise growth during homomorphic operations, Inco has implemented efficient Bootstrapping technology to refresh ciphertexts, maintain decryption capability, and perform complex computations simultaneously.
Fhenix focuses on providing robust infrastructure for privacy-preserving applications, utilizing FHE to offer end-to-end encryption solutions to protect user data. Fhenix’s platform aims to support a wide range of applications from secure messaging to privacy-preserving financial transactions, ensuring data privacy throughout all computing processes.
End-to-End Encryption: Fhenix ensures that data remains encrypted throughout the entire process from input to processing and storage. This is achieved through a combination of FHE and Secure Multi-Party Computation (SMPC) technologies.
Efficient Key Management: Fhenix integrates advanced key management systems for secure key distribution and rotation, critical for maintaining long-term security in an FHE environment.
Scalability: The platform utilizes optimized homomorphic operations and parallel processing to efficiently handle large-scale computations, addressing one of the main challenges of FHE.
Co-Processors: Fhenix has also pioneered dedicated co-processors designed to accelerate FHE computations. These co-processors handle the intensive mathematical operations required by FHE, significantly improving the performance and scalability of privacy-preserving applications.
Zama is a leader in the FHE field, known for its development of the fhEVM solution. This solution allows Ethereum Virtual Machine (EVM) computations to be performed in a fully homomorphic environment, ensuring privacy at the execution level for any L1/L2 projects built using this library.
fhEVM Solution: Zama’s fhEVM solution integrates FHE with the Ethereum Virtual Machine, enabling encrypted smart contract execution within the Ethereum ecosystem for confidential transactions and computations.
Concrete Library: Zama’s Concrete library is a Rust compiler for TFHE (a variant of FHE). This library provides a high-performance implementation of homomorphic encryption schemes, making encrypted computations more efficient.
Interoperability: Zama is committed to creating solutions that seamlessly collaborate with existing blockchain infrastructures. This includes supporting various cryptographic primitives and protocols to ensure broad compatibility and ease of integration.
FHE in Crypto and AI Infra: The intersection of cryptography and artificial intelligence is gaining momentum. Innovative models and datasets will be driven by open collaboration among multiple stakeholders, with data being a crucial part of this cooperation pipeline. FHE plays a key role in training models without revealing the underlying datasets, potentially unlocking the monetization of datasets for enhanced open collaboration between data owners.
FHE Enhancing Privacy-Preserving Machine Learning (PPML): FHE enables sensitive data like medical records, financial information, or personal identifiers to be encrypted before input into ML models, ensuring data confidentiality even in compromised computing environments.
The Future Driven by FHE: While the industry debates whether a unified FHE solution will dominate, the diversity of requirements for different applications may necessitate specialized solutions optimized for specific tasks. Interoperability between schemes could be a practical approach to flexibly address diverse computing needs while leveraging the strengths of various solutions.
Availability of FHE: The availability of FHE is closely linked to reducing computational overhead, improving benchmark testing standards, and advancing the development of dedicated hardware. Despite challenges in interoperability, computational costs, and hardware support, the potential of FHE in blockchain, privacy-preserving machine learning, and broader Web3 applications cannot be overlooked.
In conclusion, FHE provides a powerful tool for data privacy protection and secure computation. While challenges remain in interoperability, computational costs, and hardware support, FHE’s potential in the fields of privacy protection and secure computation in the future is promising.
