Interview: Manta Co-Founder Discusses the Role of Privacy in DeFi

As an early-stage project, Manta Network considers itself a “privacy-preserving” DeFi stack that leverages zero knowledge proofs. The protocol aims to facilitate transactions with guaranteed anonymity and a degree of cross-chain interoperability.

Manta has recently gained attention after the announcement of a $1.1 million fundraising round, led by Polychain Capital. The funds will be used to build MantaSwap: a decentralized exchange (dex) aimed at safeguarding user privacy.

Shumo Chu is the CEO and co-founder of Manta Network. He recently sat down with The Tokenist to discuss the details behind Manta Network — how the protocol facilitates anonymity, the technology it supports, Manta’s role in DeFi, and much more.

Tokenist: Hello Shumo, thanks so much for taking the time to chat with us! We’re thrilled to learn about the details behind Manta Network – an emerging project that has quickly gained momentum. We think Manta is doing a lot of really cool things when it comes to user transactions and privacy. But before we dive into that, perhaps we should start with a general introduction. What exactly is the Manta Network?

Shumo: Manta Network is the first privacy preserving DeFi stack built on Substrate and powered by zero knowledge proofs. Manta includes a decentralized anonymous payment scheme that supports DOT and parachain assets (for example, aUSD) and a decentralized anonymous exchange scheme that is a AMM that is built by zkSNARK.

Tokenist: What types of digital assets does the Manta Network support? Will it eventually support the anonymous ownership and transfer of synthetic assets? 

Shumo: Manta will support DOT and standard parachain assets. We have seen many efforts to bring ERC20 assets to Polkadot, if these assets can be wrapped as a parachain standard token, Manta can support them as well. We are also looking at synthetic assets, developing a scheme for a private synthetic assets is interesting and may require some further work.

Tokenist: How does Manta make a transaction anonymous?

Shumo: Very nice question! Making transactions anonymous is hard actually, Manta leverages zkSNARKs to be able to verify computation without leaking any information about the input and computation itself. However, simply applying zkSNARK to existing consensus and AMM would not make it private. You need to design a privacy preserving protocol that combines many things together, for example, accumulators, asymmetric encryption, cryptographic hash functions, and of course, zkSNARKs. The best way to learn how Manta make transactions anonymous can be found in Manta Workflow and Manta Whitepaper.

Tokenist: Why is anonymity important in DeFi? How does anonymity protect users, and how does anonymity separate DeFi from the traditional financial realm?

Shumo: Privacy is extremely important in not just DeFi, but blockchain in general. We recently conducted a survey of 400+ respondents to learn more about their sentiment towards the state of on-chain privacy–essentially, wallet address pseudonymity. About 75% of them said that they have either hesitated or altogether avoided making a transaction in the past due to their concerns over privacy. In other words, on-chain privacy isn’t just something that people are thinking about, it’s actually impacting user behavior. Our team can share many stories about it–generating new wallet addresses to receive tokens, sending out transactions from exchanges to obfuscate our wallet content, etc.

With regards to DeFi vs. traditional finance, I don’t think privacy separates the two; I think it actually provides similarity–a user experience that most people are familiar with when dealing with TradFi. For example, if you want to view your transaction history and account balance in your debit account, you have to log into your bank account. You enter your unique username, your password, maybe do an email authentication or other form of 2FA, and then you might even need to answer some questions if you’re logging in from a new IP address. You go through all of these security checks just to prove that you’re you, and only when you successfully pass everything, you can view your information. It isn’t like that in blockchain. 90% of the respondents in our survey admitted to peeking into other people’s wallet addresses to view their assets and transactions. We all do it. It’s a huge, unaddressed invasion of privacy that needs to be solved. I think the difference between privacy in DeFi and TradFi, though, is that there is no authority that can circumvent an individual’s privacy in DeFi. In TradFi, the bank you bank with can circumvent your privacy. A bank employee can access your information, view your data, etc.; in DeFi, on the Manta Network, no one has access to your data except yourself. 

Tokenist: Some anonymous payment methods currently exist in the DeFi space, such as Zcash. What does the Manta Network bring that’s otherwise lacking in DeFi?

Shumo: With respect to the category of layer-1 privacy solutions, the greatest value that Manta Network wants to bring to the table is interoperability. Manta Network intends to operate as a BYOT (bring your own tokens) system. Users can transact or swap DOT privately, for example. There is no need to change DOT into a different asset before using it.

Tokenist: Another attractive aspect to DeFi’s landscape (and blockchain in general) is increased access to financial services. Yet many centralized exchanges are becoming increasingly regulated, restricting their user base and ultimately creating barriers to entry. Will true anonymity help to make DeFi more accessible?

Shumo: Regulation and compliance seem to have a bad reputation in blockchain, but the reality is, it is a critical piece of the ecosystem welcomed by the community. In fact, in our survey, many users prefer to use certain exchanges due to the regulatory aspects of those exchanges, and the safety and security that comes with it. 

DeFi takes an automated approach to access. When it comes to mixing DeFi with compliance, what we see is a financial system where good actors (e.g., average Joes like you and me) are able to access financial products and systems instantaneously, while bad actors (terrorists, blackmailers, etc.) can be gated. The responsibility for gating bad actors is something that is currently tricky to solve for in a decentralized world, but it doesn’t mean that we should turn a blind eye to the consequences of allowing such activity by shrugging it off as “free market” or “decentralization.” DeFi makes access to financial tools more accessible, and anonymity enables access to those tools without worry about unauthorized monitoring. Compliance provides the sanctity of those financial tools. We believe there is a world where DeFi, anonymity, and compliance can all exist.

Tokenist: Is Manta’s protocol being built to incorporate regulatory compliance? Is there a way to maintain anonymity to protect users from bad actors, while also maintaining regulatory compliance, such as KYC/AML for example (assuming regulatory authorities are not “bad actors”)?

Shumo: I think I alluded to this in the previous answer. We are taking a proactive stance in exploring all the various ways of protecting our community and network. We want Manta Network to be a plug-and-play privacy tool to enhance the user experience of good actors when executing on-chain actions. As an early stage project, we are still in the exploratory phase of figuring out exactly how to incorporate compliance, so we don’t have a clear answer on exactly how we are executing it at this time, but yes we are exploring it. 

Tokenist: Both Manta’s Decentralized Anonymous Exchange (DAX) protocol and Decentralized Anonymous Payment (DAP) protocol use Zero-Knowledge Succinct Non-Interactive Argument of Knowledge (zk-SNARK) proofs. What are the primary benefits of non-interactive zero-knowledge proofs, as opposed to interactive zero-knowledge proofs? The common sense answer would suggest something along the lines of quicker transaction speeds, i.e. greater transaction processing power, and therefore the ability to scale at large. Is that the case?

Shumo: For interactive proof systems, the prover and the verifier need to interact with each other in several rounds before the verifier can verify the proof. For blockchain applications, this brings several problems: 

1. Whenever a verifier, e.g. the consensus node need to verify the proof, the client has to be always online
2. There has to be separate communication channels between the client and every single consensus node

For non-interactive proof, you only need to prove it once and then the proof is publicly  verifiable. So for most blockchain applications, you really need non-interactive proofs, this will lead to both better usability and better scalability. Fortunately, non-interactive proof is not that hard to achieve, you can apply the so-called Fiat-Shamir transformation to convert interactive proof systems to non-interactive proof systems almost for free.

Tokenist: Is it possible for a false proof to appear as valid to a verifier? What ensures this does not happen?

Shumo: No. More precisely with negligible probability, for example, less than 2^{-128}. 

This is the cryptographic guarantee that the underlying zkSNARK provides. 

Tokenist: When a Manta-powered zk-SNARK transaction occurs, is certain meta-data recorded and publicly accessible (date / time of transaction, etc.)? If so, which data?

Shumo: There is indeed meta-data recorded, for example, the rough time that the transaction occurred in the blockchain. However, to the first approximation, the entire transaction is encrypted, leaking the time information is not that critical most cases (there are still cases that leaking time information could cause trouble though).

Tokenist: Does the future of quantum computing pose a threat to zk-SNARK — and encryption in general?

Shumo: There is some on-going research on post-quantum zkSNARKs. Actually, our co-founder Zhenfei Zhang is a leader in this research field. My understanding is post-quantum encryption is relatively easy, people know how to do this now. Efficient post-quantum zkSNARK is still quite hard now. But I have confidence that the community will figure this out before quantum computers become the real threat. If you talk to any quantum computing people in private, they will tell you that quantum computers still have a long long way to go, despite that they will say some politically correct things in the media to promote the field. I can totally understand this, you need a little bit of hype to move the field forward.

Tokenist: Can Manta’s code base be audited by third parties?

Shumo: Yes, we will ask the top security audit firm to do third party auditing on Manta code. Also, we will do security auditing and fuzzing internally as well.

Join our Telegram group and stay connected to all things crypto, DeFi, and finance.