Guy Itzhaki on encrypted smart contracts and private DeFi

0:00I'm Ashton Addison from the Crypto Coin Show and today on Blockchain Interviews with Guy Itzkovich, co-founder and CEO of Phénix here to discuss FHE, which is fully homomorphic encryption, what that means, why it's important for smart con- smart contracts and DeFi, and how better encryption can unlock private DeFi, hopefully more institutional adoption, and greater mainstream

0:26adoption of crypto and decentralized finance. Guy, welcome to the show and thank you for taking the time. Thanks Ashton for inviting me. Super happy to be here. There is a lot to cover. We talked You mentioned FHE, you mentioned, you know, Phénix the product. We're going to have definitely an interesting talk. Definitely. I am in for anything that helps

0:49DeFi and crypto overall get into the hands of more players and the big money, they generally they want encryption, they want to ensure that it's secure or more secure than the traditional finance as well. I think that's important for the next steps. I'm excited to hear how Phénix plays a role in that. I would love to start out with more of a high-level on what you and the team have

1:11built at Phénix and how it relates to encryption and then we can dive into the details. For sure. So, if I need to portray Phénix in like one word or one sentence, it is that we're building the fastest privacy infrastructure for blockchain powered by fully homomorphic encryption. And I know fully homomorphic encryption is kind of a mouthful and it sounds very complex.

1:32We'll we'll explain what it is. We'll keep it high-level. But, just a little bit of like background about myself and also about the company, I spent a big amount of my career, or as I say in my previous life, at Intel. I was leading the team that was building the trusted execution environments of the world. So, a lot of the SGX project that you see out there

1:55on the blockchain space that you see these are kind of my kids because I invested a lot of time in just making sure that TEEs are out there in the world and I love TEEs. I think they're super valuable. Around 2020, I decided to take a little bit of a leave of absence from privacy and confidentiality and I was a technology advisor for the Olympic

2:15Games, which is kind of weird, but still had a pleasure of working on the Tokyo Games and the Beijing Games during the COVID period and then moved back to the space and I was managing the team that was building the hardware acceleration for fully homomorphic encryption, building a business around that and I figured out a trajectory that FHE is actually

2:35gaining in. FHE has been out there for many, many years. It was considered as like this magical cryptography that enables you to compute over encrypted data. It was always considered as very complex and very slow. We'll tackle these challenges in a minute. And then a couple of years ago decided to leave corporate America and join the other guy, Guy Zyskind, who built

3:01Secret Networks, one of the I [clears throat] would say earlier pioneers of privacy. They were quite strong in the Cosmos ecosystem. And the vision for Phénix when we started Phénix was actually to solve the privacy challenges that we see on blockchain today. I'm 100% with you that like if you want to bring real money, if you want to bring the large player in,

3:24if you want to see like massive adoption there needs to be some level of maturity that is coming into this world that we're adding. And that maturity includes privacy, but not just. It also includes compliance and the ability to do both of them because many people when they think about privacy they say, isn't privacy just for like tax evasion or anything

3:44like that? And it's really not. Like the goal here is to bring privacy in a way that doesn't impact the business part of it and actually maintains business to maintain compliant. And we're 2 and 1/2 years into the journey. It's been very exciting. Mhm. Yeah, that is really exciting. That's quite the journey. And it sounds it sounds exciting

4:08to me. Security for the Olympics. And then diving back onto the web 3 side, the dark side. Once people from Intel go into crypto, there's no going back. There is no going back.

So, welcome to the dark side. And with this fully morphic encryption, FHE, I feel like it's like smart contracts in a sense that it almost was around before

4:31crypto and for other purposes and just encryption has been a discussion since like the '90s. But can you explain what is encryption versus a FHE and why it's better? For sure. So, we take a step back for a second. When we think about privacy, we always think about like it is one thing. But, there are actually I would say different flavors of what

4:56privacy means. The first part that many people relate to is anonymity or the ability to hide the identity of the person that does the transaction. This I wouldn't call it necessarily encryption, but more of a proving mechanism where you actually prove that the transaction has been executed without sharing the identity of who did the transaction. And

5:21this is where zero knowledge proof plays a significant role. We see projects like Railgun and others that enables the ability to achieve anonymity, dark pools and other mechanisms. Where we come into place is the ability to operate over encrypted data. So, the reality is that today and I don't think I hope people know it that everything

5:46on blockchain is transparent. You open a block explorer, you can literally see like everything that's happened from from day one. Mhm. And we think it's a problem. It's not just us that think it's a problem. There are other projects that are trying to do the same, but we also get the same feedback from institutionals and DeFi projects that want to maintain their

6:04strategy confidential or whales that want to maintain their balance or institutionals that want to do payments and don't necessarily want people to see how much was paid from one one user to another. So, this is where you would actually want to encrypt the data and compute over that encrypted data in a way that nobody from the outside can see what the

6:26content of the transaction is, but you can still operate over that over that encrypted data. There are a couple of solutions that are trying to solve it. As we said, zero-knowledge proofs are probably not in that realm. They're more in the anonymity part and we're talking about confidentiality. Trusted execution environments are trying to tackle this ability to compute

6:50over encrypted data. Trusted execution environments, as I said, I think they're great. They have a limitation of trust. You need to trust the hardware. And a couple of months ago, we actually saw that with not a lot of effort, you can actually build hardware that can listen to the memory and basically see the content of what's being computed inside the

7:11trusted execution environment. So, they're amazing for specific applications, but in a world where you have a decentralized ecosystem and people can potentially get their hands on the hardware, they're probably not the best fit. So, how do we solve it? We solve it through math. And when we talk about math, FHE is really where we're leaning to. So,

7:35fully homomorphic encryption, it's not a generic encryption. It's a specialized type of encryption. And the word homomorphic comes from the world the word of like maintaining the ability to compute over the data even though the data is encrypted. And the way we do it is that we encrypt the data in a special way. We still maintain the mathematical

7:59structure of the data that we encrypt. But now we can compute over that encrypted data. Which sounds a little bit like magic because when we were taught in school, we always said like if you encrypt something it's kind of meaningless, right? You need to decrypt it if you want to compute over that. FHE changes that dynamic. And that is really the core underlying

8:19cryptography that we use in Phoenix. Mhm. That's really cool. And yeah, I've always understood that once you encrypt something you really have to decrypt it to do anything with it. But to [clears throat] combine that you know, those ZK proofs and other technologies to actually use it while it's encrypted is ideal. So, you know, I've there's been a bit of

8:41a privacy narrative throughout the second half of 2025 moving into 2026 still with the rise of Zcash's project, Zcash, Monero, Ethereum, Vitalik talking about privacy. Why do you think this topic is mattering more than ever? And does it have to do with the rise of AI?

9:07so, personally I'm of course very very happy that the rise of privacy is here. I will say when we started it like 2 and 1/2 3 years ago, we had to go and convince people that privacy was important when you know, people said, "Yeah, I don't care. I use Gmail and everybody knows everything about me." And we had to tell them, "But there is a big difference between like

9:29what you as a user perceive as privacy and really what the market needs in terms of privacy." For me, there are a couple of reasons why I see that actually happening. I would say the first one is the stablecoin narrative. I think stablecoin brought a product-market fit to crypto. And with stablecoin coming in, now when you have a product-market fit,

9:55now you start to figure out, okay, so what can I do with it? I can start doing payments. And then privacy matters. When we talked in the past about use cases like meme coins and NFTs, it was hard to explain what privacy really matters. But when you start talking to like financial, this is where it becomes just, okay, we need to solve it and the next problem, which is again

10:17privacy and compliance. So for me stablecoin is of course one of them. The other part, which I would say is also correlated to the stablecoin, is the institutional demand. As we see more and more institutionals coming in, they're like so JP Morgan Chase and Robinhood and others, they come with like privacy mindset that there's, you know, has to be

10:38part of the victim. And I'll always iterate it. We talk about privacy, but it needs to be in a way that allows them to maintain compliance. Because what do institutionals want? They still want to make sure that there is KYC. They still want to make sure that there is AML. So you need to offer a solution which is flexible enough to give them the privacy

10:59levels that they need, but that does not now open the door for the regulator to come and tell them, hey guys, you know, this is not what we need. So I would say it's it's it's it's the stablecoin, the institutionals, and then I do agree with the rise of AI. Even though I have to say I still think that it's a bit early because I think we're still trying to figure out now the

11:23AI part of it and we see a lot of like agent-to-agent payments, etc. Once there is like again like very clear product-market fit for XPRT for that matter, then I would say we will probably also see privacy as the next order. Definitely. Yeah, stable coins brought that product market fit and I feel like it's a natural transition that from people institutions using stable coins

11:51for payments to AI agents using stable coins for payments and the current narrative now in the last couple of months seems to be that crypto payments and stable coin payments they were meant for AI agents not not for people. Yeah, I mean we definitely see the same trend in terms of the amount of project that are now starting to look into like

12:15agent payments. We did an implementation of private XPRT about 3 months ago. So people that you know you want to try it, people can can go and check it out on our website. But for me we'll see like progression that's happening because even though you're right that there was a lot of conversations on privacy happening since I would say Q3 of 2025

12:42from that these conversations to a point where you actually see production implementations out there in the market, there is a process and this is one of the things that we're now working quite a lot is A to make sure that the infrastructure is ready for the needs of the customer and how do we build this infrastructure? So that's maybe like one of the things that is

13:05important to mention. When we started Phoenix, we started Phoenix as an L2. That was like the original assumption that the market needs another L2. That was about 2 and 1/2 years ago. Very quickly we came into this realization that the market doesn't need another L2 and we actually want to meet developers where they're at on the chains that they're currently operating.

13:28So, we switched from an L2 architecture to what's called a coprocessor architecture. An easy way to say is privacy as a service. Mhm. So, our product now taps into EVM chains. So, Ethereum, Arbitrum, Base, and we can add more as needed. And it enables developers operating over these chains to now introduce confidentiality into their smart contracts. So, instead of

13:55telling them, "Hey guys, you want to use our product, you know, come work on our L2." Now you're actually saying, "Okay, this is very easy for you guys to start integrating confidentiality into your smart contracts. You just import the libraries, and from that moment onwards, you can start encrypting decrypt the assets in your smart contracts."

14:13So, that is like one example of how you reduce the entry barrier for developers to start implementing privacy. That is one part. The second part is you want to maintain Solidity support. You don't want to have them to become cryptographers and really start to like understand FHE because if I have to count on developers to know FHE, we can close the shop and

14:37go back to the Olympic Games. So, Solidity make it very very simple. And the last part is performance. If there is interest, we can explain like, you know, the performance improvement, which is quite astonishing. Yeah, no, I think that transition makes total total sense. We need to upgrade the for example, the DeFi applications that have a network effect that are

15:00already on Ethereum or the L2s or, you know, the Binance Smart Chain or some of these other chains. So, for example, if there was Aave or Uniswap as a DEX, they wanted to implement compliant private DeFi. They want to add more privacy. They would be able to use this technology fairly simply to upgrade everything and make it more private

15:26but still be compliant. 100% I would say the mechanics of the actual integration are very very simple. As I said, it's solidity, it's the same tools, it's something that you know is they're familiar with. So, implementing it makes a lot of sense. I would say where the challenge is just figuring out the right mindset and the design itself. Because if you're starting to move into

15:53data which is not necessarily public, you need to figure out that you design it in a way that does not expose the data from other mechanics. So, if you for example, you have a pool and you extract an amount, if the pool is public and the amount that you extract is private, it's very easy to figure out how much you've extract. So, you need to do like a bit of a different thought

16:17process on the actual design. But, the beauty is that they can actually integrate this very very quickly. And what it would give them is the ability to operate over encrypted data but still do KYC, AML if needed, as well as in certain conditions what we call KYT, know your transactions. Because sometimes we also get this request, okay. There are certain

16:43scenarios where maybe the regulator or an auditor would come and say, "Hey, I actually want to see the content of this encrypted transaction." Now, we don't have access to the key, so we can do that. Of course, that's like where where we're operating. But, we can help the smart contract developers to basically define what are the conditions that given that they're fulfilled would

17:07actually enable a third party to see the data. And that really helps also close that gap of meeting the compliant needs. Definitely. So, you guys were way ahead and with the other guy and in Secret Network been focused on privacy for many years now. With this multiple years, now the narrative is catching up. People are realizing in the last few quarters privacy is

17:33needed. So, what exact stage is Phoenix at with actually implementing this into some of these huge DeFi apps like the ones I mentioned or people that are building is it deployable and usable right now? So, we are on testnet on Ethereum, Arbitrum, and Base and we're looking into expanding into additional chains. So, that is like step number one is just

17:59making sure that developers can access the coprocessor capabilities on these chains. Mhm. The next step is now getting protocols and projects to implement and enhance. So, we tackle this with multiple funds. One is builders. So, we have today somewhere I would say like between eight projects that are actively building on us and we're starting now a buildathon and

18:26other mechanisms that would increase the amount of builders that are actually building on us. And we see really impressive projects that are being built. So, a private lending protocol, confidential transactions, confidential payments. All of these are projects that are starting their journey and are already building on us. So, that is like the first part, building the

18:48the beginning of the tier. The second part is reaching out to DeFi protocols that are out there and exposing the capabilities to them and having them integrate that. And also there we are now in discussions with a couple of them that are tackling it from multiple fronts, whether that may be payments or dark pools or OTCs. They all understand the value and are

19:13now looking into implement it. And then you have the last year, which is the large-scale institutionals of the world. And also there are active conversations happening, which we're very very excited because we are talking to the really largest player out there. And it is really heartwarming to see the fact that now finally after many many

19:39years of conversations institutionals are actually looking into moving into Web3. Mhm. Yeah, I feel like it's similar to you know, watch what they watch what they do not what they say in a sense that these institutions, you know, for a long time they were sort of bashing Bitcoin while accumulating it. And on the privacy aspect, the narrative was that, you

20:03know, it's for criminals or you know, some negative connotations. Well, meanwhile, they're not actually going to come in unless it is private, compliant, but private. And I and it's that technology is going to be the foundation that they want it to be more private. It's really very transparent on all these chains right now. Yeah, it's it's always, you know,

20:24amazing that even now when we have conversations, sometimes the reaction is, " privacy isn't that really for criminals and tax evasion." And we tell them it's not it's not really the case. We've progressed so much more and we have the right infrastructure on all fronts. It is literally Now that there is product market fit, stablecoins, payments, just come and

20:48move. We'll help you we'll help you integrate. I was in Consensus Hong Kong about 3 weeks ago or listened to a couple of the institutionals play and even Google for that matter are now accepting crypto payments, but are also enabling payments for YouTubers with crypto. So, you definitely start seeing the proliferation of the benefits that we

21:13can get out of blockchains and to the more standard web two entities. Definitely. There's There's always that last mile and the endpoint where you're in an encrypted ecosystem, but then you got to pay your bills in fiat and like you need to come out and then you know, that the whole system needs to be encrypted together for at least with the crypto natives that have wanted to

21:40pay with stablecoins. It's not It's It's not there yet to be able to live your life. Yeah, I always say that like the best approach is to be very very practical. I don't think we will ever operate in like a fully encrypted ecosystem. I think you would probably need to define what it is that you want to protect and whatnot and just operate around that

22:06And make it very very easy to deploy. So, as an example, one of the showcases that we released a couple of months ago was the ability for stablecoin issuers to issue stablecoin that has a dual mode. It has a private mode and it has a public mode. The other option is to say, let them issue just a private stablecoin. But, if you just have a private stablecoin, it's

22:31very hard to integrate that private stablecoin into yield-bearing DeFi applications because the likes of Aave in that regards, they can't use the private stablecoin yet. Mhm. They will in the future, but they cannot use it yet. So, issuing a stablecoin that has a dual mode, a private and a public, which enables the users to decide, you know what? Now I actually want to make a

22:56transfer, I want to hold a balance, and in that case I want it to be private. That's fine, that's what I need. But, if I actually want to start using it for like a yield bearing mechanism and the entity that I'm going to work with does not support that, then I can switch it into the public mode, put it there, they don't need to change anything, and I can

23:16start getting yield as well. And this is really a progression that we'll see happening throughout time until we reach the end state, which is, you know, we just all converge on a solution that meets like both privacy as well as like over the private state. That's really cool. Do you think we could see that with USDC and USDT to have private versions or

23:42at least

We definitely hope so and we are reaching out to stablecoin issuers for exactly that. That's

That actual integration, I have to say, is really simple. Like, if you if people in the past thought that, you know, starting to now to converge into like a private stablecoin is very complex, it's not really. We can like again, four lines of code, you can now

24:09issue a new stablecoin that is either fully private or has this dual mode. So, it's just starting to build the momentum and the demand. That's really cool. So, you mentioned that you're on testnet, developers can start playing where where they're at, you know, whatever chain, all the ones that you have already integrated in the languages that they already use, making

24:31it easy. What's the next step of gaining more adoption on the building side and on the implementation side? So, we are on testnet and we are planning to move into mainnet towards the end of the year. That, of course, means a lot of work on our part. Moving from testnet to mainnet has, you know, significant implications because we want to make sure that

24:56our product is secure, safe, and fast. That is one part on the engineering front. Regarding the builders path, so we have always new versions and iterations of our SDK. Again, making it significantly easier for developers as well as reducing the impact on the user experience at the end of the day. So, we'll we'll keep on seeing like new iteration. We actually have a new SDK

25:23coming up at the end of this week, and I would very encourage developers to look into that starting next week. It's it's it is a major change and improvements in the SDK. And then there is, as I said, the demand for it, and that is where we onboard developers, builders, as well as building a few initiatives in-house, which we can probably share in the next

25:49In the next time we're invited here. A couple of things that are super super exciting for us. Very cool. I'm looking forward to it, and would love to have you back as this develops out and gets integrated more into the industry. What's the best way for developers to get started and for other people that are they're becoming developers, or

26:10everyone's becoming a developer now with AI. It's making it a lot easier. But for people to test it out and play with it. So, all the information is available on Phoenix.io. I think you're spot on with the integration of LLMs. We see people vibe coding very very quickly. I wouldn't encourage people to go into production based on a product that

26:32was vibe coded. That's probably maybe not the best approach for now, maybe in the future. But for ramping up, definitely we've seen just about a week ago, we did the an open house together with Arbitrum in New York, and projects that are building on Arbitrum are now starting to look into integrating a coffee as well and the feedback was very very simple

26:56very It was great. They were able to integrate coffee into their product very very quickly. So, it's not complex. So, either use your favorite LM tool or just read the documentation on phoenix.io. We really like to work with developers. I have an amazing dev rel team. Anybody who would need support, they can reach out to us and the dev rel team will

27:20will definitely help. Sounds great, Guy. I appreciate your insights into private, compliant DeFi, payments, and crypto overall. This can this can really upgrade the whole industry. I'm looking forward to following along as you get closer to the mainnet. Would love to have you back on. I can share a link to the socials, the site, phoenix.io in the show notes below,

27:46and really appreciate your time. Thank you so much. Would love to follow up again in the near future. Thanks for inviting me. It was a was a real pleasure. If at some point you want to talk about technology and Olympics, I'm here as well. But, you know, just just to close this, I do think privacy is something which is transformational for crypto, and I do

28:09think that like bringing this into the market would make a lot of good for us all, not just the DeFi projects.