MindChats EP17 Recap: Meet ZAMA’s CEO

Mind Network
3 min readJun 24, 2024


This special episode of MindChats centers around Fully Homomorphic Encryption (FHE) and its intersections with AI. Hear about the efforts of Zama and Mind Network to enable more secure, fair, and efficient AI applications in Web3. Moderated by our host Christian, in this session, Rand Hindi, CEO of Zama, shared his opinion on the potential of FHE and AI with Mason, Co-founder of Mind Network and Ashely, Head of Research of Mind Network.

Recording MindChats EP17 — Meet ZAMA’s CEO: https://x.com/i/spaces/1YqKDgogwwDxV

Key Highlights

Fully Homomorphic Encryption (FHE) and Its Significance in AI

“You need FHE if you want to do anything that requires confidentiality.” -Rand.

Fully Homomorphic Encryption (FHE) enables computations on encrypted data without decryption. For instance, when using AI services like ChatGPT, FHE allows users to send encrypted queries to platforms like OpenAI, ensuring privacy throughout processing. The encrypted data remains secure, and users can receive encrypted responses that they decrypt locally. FHE is crucial for achieving end-to-end encryption in online services beyond messaging, offering significant potential in cloud SaaS products and decentralized blockchain protocols where data confidentiality is essential.

Rand elaborates on the main issues with centralized AI, including trustworthiness and the threat to privacy. In centralized AI, users do not know if the computation model is authentic or if it could be manipulated remotely. Decentralized AI offers public verifiability, which ensures the integrity and honesty of computations.

Centralized models create single points of failure that can be exploited, unlike decentralized systems which offer better protection but at the cost of open data. This is where FHE becomes critical to ensure privacy without compromising decentralization.

The difference between FHE, ZKP and MPC

Rand breaks down various cryptographic techniques — FHE, Zero Knowledge Proofs (ZKP), and Multi-Party Computation (MPC). Each has its strengths and limitations. FHE allows encrypted computations, ZKP provides evidence of computation without disclosing values, and MPC enables secure multi-party computations. These techniques can be combined to provide robust solutions in decentralized settings.

Zama’s Contributions

Rand introduces Zama’s suite of tools like Concrete ML SDK and fhEVM, which aim to simplify the development of encrypted applications. Concrete ML allows developers to use Python for machine learning on encrypted datasets without needing deep cryptographic knowledge.

Applications and Use Cases of FHE

Rand and Mason illustrate practical applications, including encrypted image processing, credit scoring, and medical data analysis. Rand envisions fully decentralized, confidential AI networks that revolutionize data privacy and utility across sectors. Mason adds that FHE contributes to equitable AI network validations and fair data transactions.

Mind Network’s FHE Validation Network and AI Subnet

Ashley underscores that the FHE Validation Network and AI Subnet use encrypted, private consensus to prevent validators from copying each other’s results. This encryption guarantees that validators conduct authentic computations, preserving the usefulness and value of AI models. Integrated within Zama’s framework, this approach plays a crucial role in identifying and ranking the most market-relevant AI models, offering valuable guidance to developers in the decentralized AI landscape.

About Mind Network

Mind Network is the first FHE Restaking Layer for PoS and AI Networks. Their framework operates as an FHE validation network, bringing secure computation and consensus to Bitcoin restaking, Decentralized AI, Bridges, EigenLayer AVS, Bittensor Subnet, and many critical PoS networks. Mind Network’s solution enables verifiable decentralized computation over encrypted data.

Mind Network is backed by Binance Labs, Hashkey, Big Brain, Chainlink, and received an Ethereum Foundation Grant for their FHE research on Ethereum.

For more insights and updates on Mind Network’s innovations, visit their website and socials:

Website | Twitter | Medium | YouTube | Github | Gitbook
Discord | Telegram



Mind Network

An FHE (Fully Homomorphic Encryption) Restaking Layer for POS and AI Networks. https://linktr.ee/mindnetwork_xyz