ZKPs and privacy pools: Why Ethereum requires privacy

Aztec Labs CEO Zac Williamon describes why privacy is not just a technological upgrade, but rather a must.

Zac is co-founder, CEO and president of Aztec Labs. layer-2 Privacy-focused network bringing Ethereum to the forefront (ETH). He earned his PhD in particle science from Oxford University and spent time at CERN before he became a crypto expert. In the blockchain industry, he is most well-known as co-inventor PLONK, which has become one of today’s widely-used zero-knowledge verification systems.

Zac recently explained to crypto.news that privacy is not just an option but something Ethereum must have in order to thrive. He explains the meaning of legitimate privacy on blockchains, privacy pools that offer privacy while also complying with regulations, and how private layer-2s can make it easier for real-world assets to be added onto the chain.

What is privacy? Does it have to do with anonymity, selective disclosing, or is there something else? 

In order to explain privacy on blockchain I divide it into three main pillars.

First there is privacy for the user, which involves hiding both the identity of the sender as well as the recipient. Data privacy is concerned with keeping the transaction amount confidential. Code privacy is the last option, which hides even the logic that’s being used on chain.

For me, the Holy Grail of privacy on blockchain is to achieve all three. This is the goal we should aim for when we want to build truly private systems.

In a general sense, I would say that privacy is the capability to exploit information imbalances on-chain. For example, I may be able to perform a trade where I have information you do not. It is important to understand that this foundational skill will be used in many basic daily interactions.

This information asymmetry is evident when voting in an election. You don’t even know what I did.

CN: Which are the main misconceptions that the crypto ecosystem has about privacy?

ZWI believe that the biggest misconceptions regarding privacy and crypto are these:

a) It’s  just about tokens and private token transfers, and;

This is currently seen as a separate area from crypto. For example, you can have NFTs and DeFi.

These are both wrong and a result of the technology immaturity that has plagued privacy solutions to date. Privacy isn’t a little niche of crypto and, I think, in the near future, crypto will all be private. 

We need to offer the privacy protections that are expected by users. If we expect crypto to move beyond its bubble, interact with real-world systems, and be more than just a technology early adopter, and even compete with TradFi and web2, we must provide them.

We have a concept we call composable security, in which, like an Ethereum Smart Contract, you can define your own rules, logic and how to conduct transactions. 

You can program your digital assets but, unlike the transparent blockchains that are available, private data is a primitive of first-class. You can mask the sender of a message or its recipients. It is possible to perform compliance checks for people who need access to sensitive information, and make sure that the information remains encrypted. 

Are you of the opinion that public blockchains should offer private options in particular authoritarian settings? How should the Ethereum community define this? “legitimate privacy”?

ZW: One of the core values that blockchains have is their neutrality and lack of permission. Anyone can code their digital assets and transact using a blockchain. So, it is not really in my right to say what is or isn’t morally imperative for a blockchain. 

Both public and private Blockchains have their place. But private blockchains have a higher value and are more practical. It’s crucial to understand what legitimate privacy is, which I believe is quite easy. 

Because I participate, it’s not easy for bad actors and criminals to take advantage of the internet for their nefarious purposes.

Tornado Cash is a good example. By using Tornado Cash you are helping the bad guys, as you increase the size of anonymity pool in which they can hide. You are not if you use privacy pools.

How does resistance to censorship fit in this context?

ZW: The network itself should be censorship-resistant. Nobody should have the ability to censor protocol-level transactions. But if I was programming a smartcontract on the network, I would have complete freedom in defining what constitutes an acceptable transaction.

Privacy is an important human right and I think people should be allowed to display themselves on-chain in private. I do not believe that users should be able to use any app however they want, particularly if it goes against the developer’s intentions or rules incorporated into the smart contracts.

CN: How do you feel about the Privacy Pools Model, which Vitalik Buterin has endorsed as an alternative to full transparency and anonymity?

ZW: I think Privacy Pools is a good first step — one of many. The technology constraints were very tight when it was developed. It was a question of how to create private transaction technology that could work on Ethereum now. The ZK technology they use is primitive and therefore limits the possibilities. It’s definitely a great starting point but not the ultimate goal.

At Aztec, we strive for full programmability. Let me give you an example. ZKPassport, a ZK company is part of our ecosystem. Modern smartphones have NFC scanners and passports are fitted with NFC chips which can be used to sign digital signatures.

ZKPassport is an application that lets you tap on your passport and receive a ZKP proving you are in possession of a valid passport. You can choose what information you want to disclose — your nationality, your date of birth, your name, whatever you decide.

Use this tech to make a DeFi-only application, for example, that can only be accessed by citizens in a particular country. The proof is done automatically, using digital signatures or ZKPs. The system is permissionless and privacy-preserving. It ensures strict compliance. 

This is a much more powerful solution than Privacy Pools. Once you’ve got full programming in your privacy networks, it will be possible to build almost an infinite number of features on top.

What design patterns, or UX advances do you believe will be critical to the mainstreaming of private transactions in your opinion?

ZW: Yeah, totally. It’s my guess that PLONK design pattern is key to UX innovation. Private transactions will not become commonplace until many technological breakthroughs are made. Private transactions are more complex than transparent ones, as you can’t broadcast sensitive information on the blockchain. The client must construct the entire transaction in a private manner.

Who pays for this complexity then? Right now, in 2025, the answer is — the application developer pays, and the user pays. App developers will have to work harder than ever before in order to create a useful application. Users are also going face more challenges. The user will have to wait for the proofs and their apps may not be able to work with web3 because of the different privacy rules.

Within Aztec, my general operating principle has been: okay, complexity in private transactions is much higher — who pays? I answered: it’s the researchers in cryptography who pay by developing better ZK technologies. We did this in 2019, when we developed the first universal ZK SNARK. The ZK-SNARK has been improved a great deal since its inception. PLONK is 250 times faster today than it was in 2019. This allows for much faster applications.

There are also language designers and engineers who work on tooling. Their job is to create a programming language that can efficiently turn programs into zero-knowledge proofs — a language where writing private smart contracts is intuitive and simple. We’ve done this with Noir, the programming language we use. You can build private, efficient apps with it without having to be an expert in cryptography.

The blockchain will understand what’s private and public. For example, it can know that the transaction sender is anonymous. This is a huge amount of effort.

You need an enormous amount of software to help developers build private apps without knowing complex, advanced cryptography. As we launch our testnets we are confident in the fact that it is now easier to develop compelling private apps.

Are you a fan of layer-2s or apps like Aztec for privacy? 

ZW: It’s a complex issue, so I believe it should be dealt with by L2s and L1s who are specialized in privacy. There are tradeoffs. Ethereum would probably not be launched today if Ethereum was private by default. The development would be much more difficult, and security would increase. 

Over time, I think that L1s will incorporate more and privacy technology. Composable security requires redesigning the blockchain from the bottom up. It’s too demanding for existing L1s because that would break their backwards-compatibility with the current ecosystems. As such, at least for the time being, I believe privacy should remain the responsibility of L2s as well as the applications built upon them.

Can we protect privacy with ZKPs only, or also network-layer security measures like mixnets and private mempools?

ZW: Yes, all of it. We also need private mempools. A blockchain that is encrypted from end to end should be the goal. When I am doing something very important in the real-world, no one should be able see my actions, except for whatever app I interact with.

Only entities required for an app’s functionality should be able to see what I am doing. No one should be spying on me if, for instance, I am paying my mortgage. Even if I interact with a DAO in a place where such work is not approved, it should be safe for me to do so.

Privacy is an important human right. But it doesn’t stop at the blockchain level. We must also have full protection at the network level.

Can you answer this question? CN: Does fragmentation within the ZK tooling ecosystem (PLONKs STARKs SNARKs etc.) represent an advantage or bottleneck in terms of maturity for the ecosystem as a whole?

A great strength. ZK Tech is in its early stages. The technologies are diverse and so are the proving systems, because we don’t know yet which will be the most effective long-term. In this area, research is updated every six months.

All technology solutions come with some trade-offs. Some tradeoffs may be more appropriate in certain situations than others. We need to experiment. We must experiment with a wide range of different ideas.

Minitel, a French network that was standardised early on in its life cycle. France had an early version of internet in the 80s. This was because the French Government built a proto information network.

People could access things like train tickets, university exam results — all kinds of services. They chose a terrible architectural design. The architecture was highly centralized. Minitel was a very centralized system. Unlike the internet today, which anyone can create a site, you needed to ask for permission from the government in order to operate an app.

Then, they stalled because they standardized on the incorrect architecture. At this point, standardizing anything ZK is way too soon. It will take a lot more research and experimentation to find out which ZK is going to last.

Fully homomorphic cryptography is another new privacy technology. What is the current status of FHE? Can you imagine the possibility that the FHE market will soon be flooded with the first full-fledged applications?

ZNIt needs more time in the oven. You should listen to experts who have no financial interest in the hype surrounding FHE. You’re too young!

It is a huge amount of work to use FHE. Yes, it is good, but for very few use cases. FHE’s current state is very comparable to ZK’s in 2010.