Bài phát biểu của Sergey Nazarov tại SmartCon 2025: Xây dựng hệ thống tài chính đảm bảo bằng mật mã

Thank you all. I truly appreciate everyone coming together. I think this event is always a landmark annual event for all of us in the Chainlink community to meet, learn from each other, and really see how we can take Chainlink to the next level. I'm genuinely excited by the wonderful mix of people I see here and at our events. There are people from the DeFi world, extreme crypto enthusiasts, and then also people from TradFi. It's an interesting combination.

I think we're still working to make it a reality, but it's gradually happening. Everyone is realizing that working together is a good thing. I also think this is an interesting time for our industry, at a unique point where it will become, in my view, mainstream now. For the longest time, I've heard about people's desire since the early days of Bitcoin, when it was about microtransactions rather than a store of value, for mainstream adoption of our industry.

That's a shared goal that our community and every other community has. My view on that has always been that it must be extremely collaborative, bringing value to all regions of the world and all different ecosystems. You don't need to remove entire regions from the world. You need to bring them into the fold. You need to have them operate on technology representing our industry, and you need to have them operate in ways our industry defines as correct. This makes certain problems impossible and makes the world work better.

I think we all know and agree that the Bitcoin whitepaper emerged from the 2008 financial crisis. There were major problems before 2008 and there have been problems since. I think it's always good for us to remember what we're really doing. Yes, we're building DeFi. Yes, we're enabling people to make an incredible amount of money in DeFi, and I'm very excited about that. But our industry has truly started to address a fundamental societal problem about how systems hold value and enable people to transact, and the failure modes of those systems.

Whether it's Lehman Brothers or Silicon Valley Bank or any other examples, even some from our own industry, those failures should not happen. The financial system should not be a set of broken promises, incorrect assumptions, and a group of people quietly deciding to defraud society for their own benefit. That, to me, is one of the core ideas of Bitcoin that started our industry. It has been a core idea for me for a long time.

Now when I think about mainstream adoption, what that means to me is solving that problem in a way that brings the current financial system along with our industry. Making the current financial system and the DeFi community mutually beneficial so their economic interests align and help each other succeed. That's what will bring our industry mainstream and enable this technology to be adopted in a way that truly powers the world rather than just a small fraction of it.

So what do I want to say about this technology and why it's better? The first idea we always talk about is cryptographic guarantees. Legal promises and paper documents are very limited in that an organization or person can make a promise to you but they don't actually have to keep it. There's no system that truly ensures they keep their promise. Whereas cryptographically and mathematically guaranteed events like smart contracts, blockchain transactions, and oracle networks are guaranteed.

This cryptographic guarantee is something I deeply believe in, and I feel many people here deeply believe in. It's a new way to guarantee human relationships where there's mathematical assurance of our relationship. It's not based on my choice or your choice or anyone else's choice, or whether we like each other, or whether we agree politically. It's simply systematizing our relationship as mathematically guaranteed. That's how it will operate regardless of anyone's wishes, choices, or viewpoints. I believe that's a better way for the world to operate.

The second important foundational idea, in my view, is smart contracts as a single source of truth. When you access Etherscan or examine blockchain transactions at an institutional level, you're looking at a single, cryptographically guaranteed source of truth. This is the exact record of what happened. This is precisely what happened with ownership, or in Chainlink's case, with data, identity, cross-chain, or other things we'll discuss.

The idea of this golden source of truth is incredibly valuable. It's what the internet and tech companies tried to achieve in the 80s and 90s, and they failed. Now our industry will truly achieve that by taking all records out of systems and transactions, out of people's hands, and placing them in a smart contract. To uphold truth against encoded truth regardless of our opinions or emotions.

In the case of institutions, regulators or society need them to meet certain standards. I've been working with smart contracts almost from the very beginning, and what I've seen is the evolution of smart contracts as a technology. Smart contracts have always been fascinating to me because they make bad behavior impossible. They make it impossible for humans to do the wrong thing if they wanted to, and they can only do the right thing because the smart contract somewhat forces that outcome.

The definition of what a smart contract is has expanded over the years. Around 2014-2015, I really saw smart contracts as being about a single chain. That single chain was executing token issuance, had a ledger, maybe doing some voting with tokens, maybe a decentralized exchange, no DeFi, no data, no cross-chain, no identity, no institutions. Just one chain, one token, your key, someone else's key, and maybe an exchange. That's what a smart contract was.

When DeFi came into the mix, you enter a world where you start needing Oracles. And in my view, this is when the definition of a smart contract changed from being focused on a single chain to being focused on Oracles and cross-chain contracts. An Oracle is a critical system driven by consensus. Chain A is a critical system driven by consensus. Chain B is a critical system driven by consensus. Other Oracles are also critical systems driven by consensus. That's how I began defining a smart contract in my mind and in the minds of some of the people I work with at Chainlink.

That's what Chainlink has done. Chainlink provided the initial Oracles for data, for many different types of data, and then for connectivity. Chainlink achieved great success powering the majority of Oracle usage globally, and the vast majority on very popular chains like Ethereum. The reason for that is Chainlink reliability. Chainlink guarantees the quality of the data, the reliability, and the uptime.

A key example I'm very proud of is the recent AWS outage that saw the vast majority of systems in our industry go down, even though they claim to be decentralized and highly reliable, including data Oracle bridges and many other things that were supposed to keep working. Guess what didn't go down? Chainlink operated completely.

This is a recurring pattern. When there are market events, when there are reliability events affecting even high-cost, high-quality tier-one centralized systems and what we thought were tier-one decentralized systems, they don't affect Chainlink. It's truly impressive to have built such a reliable system with an expanding scope. That's what I'll talk about today.

The scope of what Chainlink does continues to expand, but what's always important is that as Chainlink does more, it remains reliable. Because people don't put their money, connect their contracts, or put value at risk in unreliable systems. Anyone else trying to do what Chainlink is doing is likely inferior, or in many cases clearly less reliable than Chainlink. Through reliability and this capability set, Chainlink has facilitated over $26 trillion in transaction value.

Most of that is currently in DeFi, but if we can achieve institutional adoption to the level I think we can, I can easily see more zeros next to that number. There's a lot more adoption waiting and a lot more growth ahead, both in DeFi, but the institutional space is a completely different story.

What's been important so far for Chainlink is that it provided various standards. It provided data standards, cross-chain standards, compliance standards, security standards that we'll share today and tomorrow. Those standards have created many core services. Those core services have created truly reliable smart contracts, truly reliable next-generation systems, truly reliable DeFi. Now there are truly reliable institutional applications. We're very proud of that.

But what we've really found is there are two problems beyond that. There are many types of services needed. You need more and more types of Oracles. Data Oracles, DeFi Oracles, bridge Oracles, identity Oracles, compliance Oracles. Now Oracles for AI interaction, and there will be more and more as contracts become more complex. You face the problem of creating more Oracles of different types.

The second problem you face is the coordination problem. So today I'm proud that we announced CCIP is basically operational and in production, available for anyone to write workflows for. This is truly a major milestone for Chainlink and the Chainlink community, because it introduces this coordination capability, essentially the ability to orchestrate multiple systems and services into a single application.

What's important to understand here is that as the definition of a smart contract expands to include multiple chains and multiple contracts per chain, and multiple different Oracle networks providing data, compliance, identity, all these different things, the second big problem honestly is how do you organize all of this into a single application? That's the problem we've identified.

What we're really trying to achieve with Chainlink Runtime Environment (CRE) is to become the unifying force in technology to simplify that. What I've seen in our industry before is that something like the EVM came along and made writing smart contracts very cheap, efficient, and simple. Previously it was a very complex process.

But now the definition of a smart contract, in my view, has expanded significantly. It now includes hundreds of chains, dozens of Oracles, different data sources, identity sources, something to provide to regulators. That's what a smart contract is now, especially an institutional smart contract. So which system will simplify creating those? It can't be a single chain because a chain doesn't do that. A chain doesn't orchestrate Oracles or create Oracles. A chain doesn't orchestrate other chains.

There needs to be a place where you run code to organize all the complexity of different chains, different Oracles, different data sources, different identity systems, all the various compliance requirements into a single piece of code. That's essentially what we're enabling with CRE, and our hope and goal is that CRE will create the next phase of our industry, similar to how the EVM created a major leap in innovation and growth when it simplified smart contract creation.

So please use it, try it, let us know what you think. We're excited to hear your feedback. It's operational, it's in production, available for writing workflows. Workflows are gradually being deployed to production and rolled out to different developers. They're being introduced at different speeds depending on their workflows and what they want to build.

Ultimately I can see CRE truly driving the next generation of our industry if it's done right. And a large part of that depends on your feedback and the Chainlink community building it together, building it the right way.

The basics of CRE and how it works are really the idea of workflows. A workflow is a common term in the enterprise and institutional world, where you basically build a workflow to manage a process and make an application work across systems. In CRE we have workflows, and those workflows use basic primitives like a chain reader, a chain writer, consensus, the ability to call APIs, manage secrets, and execute multiple things.

What it allows you to do is instead of taking months or weeks to build a solution that requires cross-chain data and identity, you can now build it in hours, just like you can build smart contracts in hours or web applications in hours. So let me show you an example. This example is about the verification of reserves.

Chainlink is already the largest provider of Proof of Reserves for stablecoins, gold coins, commodity coins. All these different types of coins that need to prove their reserves.

Now someone can write their own Proof of Reserves proof and basically launch their own Oracle Proof of Reserves in CRE, and attach it to a stablecoin. They can implement something called securement. This essentially prevents creating any more stablecoins unless reserves are proven. The Proof of Reserves step in the CRE workflow checks the custodian API, returns an on-chain result.

The securement contract uses that result to verify and approve or deny minting additional stablecoins. In the first part of this workflow, you have an issuer wanting to deploy securement and Proof of Reserves. They don't want to accidentally mint hundreds of trillions of dollars in coins. They want to make sure that doesn't happen because it's not good for the coin or our industry.

They prevent that by deploying this code. With just one engineer in a few hours, in the next part of the workflow, they can deploy Chainlink's automated compliance system. The ability to select policies the coin must follow. The ability to choose what the coin can and cannot do according to issuer policies is truly impressive and valuable. It creates the first real compliance layer on-chain.

This is one of the big barriers to institutional adoption. From my perspective, our goal is to do three things: develop DeFi into an industry even bigger and more successful by providing reliable and increasingly better infrastructure for DeFi to innovate and build great things.

Second, bringing institutions on-chain, and that's where automated compliance comes in. And third, making it easy and simple for DeFi to grow, and for both to work together to the point where it becomes a no-brainer. That it becomes a no-brainer to initiate an instance and deploy your DeFi protocol in a way that it can receive institutional capital because it can comply through CRE.

And the institution deploying CRE because it can use DeFi or work with other institutions. Now the automated compliance tool has verified policy conditions and will do so by examining all the different information, examining identity, the policies, and reporting back to the system.

Ultimately what I see happening with this is you'll have institutional CRE users and DeFi CRE users, and eventually they'll be able to trade, and you'll be able to see a massive amount of institutional capital flowing into DeFi because that capital can comply with the requirements to use DeFi, which currently it cannot do without this.

The next part of the flow is CCIP. We've minted a coin that we know is backed correctly. We've verified it based on our compliance policies. And now we'll transfer it across chains. All this complexity of proving reserves, complying with policies, and transferring a coin across chains is organized and managed through a single piece of code running in CRE.

To make something like this work previously, you would need 5 to 10 different pieces of code doing different things, possibly in different languages in different places. And you wouldn't be able to show it to anyone because it's just private code you're running somewhere on a server. And that code is actually the critical code defining how this contract operates.

Now you have a single place where all this complexity around how to interact with systems, interact cross-chain, interact with compliance, interact with identity all comes together. This is really intended to, similar to how the EVM increased the efficiency of creating smart contracts, this will increase the efficiency of creating these advanced institutional smart contracts.

Where once again we believe new capital flows will come into our industry. We've already seen a lot of CRE usage by institutions. Today we announced between the Central Bank of Brazil and the Central Bank of Hong Kong, where they used both CCIP and CRE. CRE was used for e-invoicing, working with ISO 20022 messages, and CCIP was used to exchange various stablecoins and central bank digital currencies.

Chainlink is really the only system that can combine all of this together, give you the ability to orchestrate all this complexity across multiple different systems, give you all the capabilities like cross-chain. And I fully expect this example will start using other Chainlink capabilities as well.

We also recently launched the Digital Transfer Agent Standard. This is a set of contracts for tokenized funds. As you can see from this diagram, CRE is very involved in how this works. What I'll really point out is you cannot build a reliable transfer agent contract without a system capable of synchronizing all of these things back to it.

And without reliable transfer agent contracts, you won't really have scalable tokenized fund capability. And without scalable tokenized fund capability, I'm not sure how our industry provides the next generation of high-quality real-world assets. CRE will create these patterns, and contract standards like the Transfer Agent Standard will be very easy to deploy.

If you want to use a transfer agent, or create one, or you are a transfer agent and suddenly realize you need to get this tokenized fund operational, instead of a 6-month process, it could be a 6-day process. You're on your way to creating a tokenized fund in a very short time. UBS is currently using this in production.

This is another thing we announced at the conference. This is really interesting because it's being used in actual production. UBS is using this for their tokenized funds, including the Transfer Agent Standard and NAVLink, the ability to feed data into transfer agents. All these more advanced things are really only possible when you have systems that can properly execute them.

Currently there really is no other system. Like Chainlink can give you data, cross-chain contract connectivity, identity, and much more coordination into a single workflow. That's just not possible anywhere else right now.

We've also been doing a lot with leading institutions using CRE and CCIP. So ANZ and Fidelity conducted a transaction using CCIP and the automated compliance tool. I fully expect something like this will be running on CRE at some point soon.

Then Kinexys, JP Morgan's private chain, announced with us and used Chainlink for a transaction between their private chain and a public chain. That's really a capability of CRE in simplifying the interaction between a tightly managed private chain and a public chain. In this case, a tokenized fund.

What you're seeing here is JP Morgan's private chain, the largest bank with many users on it, can now transact with public chains. This is a great example of what I mean when talking about the benefits of DeFi. Because basically, chains here have a lot of institutional-grade components, but in a big way they're part of the DeFi movement.

This is what you'll see more and more. From the acceptance of CRE and the success of Chainlink standards, you'll see institutional transactions, but then you'll see transactions between institutions and DeFi. Honestly, the more value brought on-chain by institutions, whether tokenized funds, stablecoins, it doesn't matter. It only brings massive benefit to our industry.

The more capital, the more assets, the more cash on-chain, the better for all of us. That's just a simple fact of life. We also used CRE in our work with Swift, which we released late last year in Beijing.

We enabled the Swift system to operate as it does, then lock assets on chains. We built an entire application around this, using Swift messages to actually lock assets, then allowing the Swift system to make payments in its traditional form and those payments unlock assets, building an application around that.

The application enabled these transactions and then returned the single source of truth from the chain. It proved two things: that Chainlink can be used very effectively, and that workflows built around existing complex systems can produce on-chain records that are the authoritative single source of truth.

That is extremely valuable because once people in the world decide that blockchain and data on blockchain from Oracles is the single source of truth, that's the point where I think we truly win. Because if you control the single source of truth and where it exists, you control how systems and the world operates.

We really need that set of systems and standards, and I think it will take the financial system and industry to a much better place where certain bad behavior is simply technically impossible.

This is truly a complex workflow and we built it with CRE late last year. We've been working on CRE for a while. It's now released to all of you. It's operational. Anyone can write a workflow and simulate it. Build workflows based on various service and system patterns. We encourage you to do that and hopefully you find it useful.

If you have feedback for us, please let us know. The deployment of these workflows is happening gradually. Right now we're also building a large ecosystem of high-quality data sources, payment service providers, transfer agents, and all kinds of service providers in the CRE ecosystem.

I think we'll also reach a point where any of you can build a service or system on CRE to sell to another developer or consumer. Some of the first participants in this ecosystem in just the past few months have been some of the world's leading data providers. S&P Global, FTSE Russell, Tradeweb, and Deutsche Börse have all in the past few months started bringing their data on-chain through Chainlink.

I think that's because of all the institutional acceptance you see in the Chainlink ecosystem, CRE, and Chainlink standards generally. I expect that to continue. I think we'll truly build a wonderful ecosystem around CRE as it matures and operates well enough. We'll really find ways to grow the data, the services, payment systems, everything easy to use in CRE as a developer or easy for you to add to CRE for your own use or for others.

If you're interested in CRE and where it will develop next, I encourage you to attend these sessions. We have sessions from product and business on the left tomorrow, and developer sessions today about nodes from our team and from Amazon Web Services. You're welcome to attend these sessions, learn more about CRE, and experiment with it as a developer.

We're very much looking forward to hearing your feedback and working with the community to make CRE a highly reliable, extremely useful, unparalleled, and efficient way to build advanced and sophisticated DeFi smart contracts.

Another feature I'm really excited about from CRE's perspective is privacy. I've been obsessed with privacy for a long time, and implementing privacy well in our industry is something we've researched extensively for many years, working with leading researchers throughout the process going back to the era of the Town Crier, where we worked with Ariel and Fan near 10 years ago.

Now I'm very happy to say that much of that great work with Ari Juels and others, like Town Crier and DECO, has finally been implemented in a unique and comprehensive approach in CRE. The security properties of CRE will come from a system that has been extensively researched and thoughtfully designed to solve key problems in how to do private smart contracts on any chain.

The key problems there are revealing the identity of users of contracts, revealing the activities of contracts, and revealing the amount and movement of value through contracts. All of those problems should be solved through the security system running on CRE.

What that system looks like at a high level, and we're announcing information about this today. We'll release the first version of this system late this year and early next year, and then we'll release more parts of it throughout next year. What we release this year will be for a small, very selective group. Early next year, it will be for more people generally, and then more general features and security features next year.

Basically, it's based on something called Chainlink DKG. There will be workshops so you can learn more, but basically it ensures that your workflow in CRE remains confidential and secret, enabling the use of AI to power those workflows with data or identity kept private. And even the outputs to blockchains can achieve a certain degree of privacy, even when blockchains don't normally provide you with that privacy.

And then all the workflow code and all the AI requests can be verified by your partner or anyone you want to allow to verify them, so you have a certain verification that the workflow and system performed everything correctly. You don't have to rely on trust assumptions of CRE or a decentralized Oracle network performing everything correctly. You're relying on trust assumptions of, in this case, a trusted execution environment.

Our approach to privacy within Chainlink and confidential computing is that you'll be able to use different privacy approaches. Like the rest of Chainlink gives you options to take this data or that data, connect this way or that way, check identity here or there, you'll be able to choose different privacy approaches from trusted execution environments to zero-knowledge proofs to fully homomorphic encryption to some combination of different approaches.

There will be more information about this relatively soon this year and then early next year. There are workshops you can attend where our research team is beginning to share this with the Chainlink community and getting feedback from peers.

We encourage you, if you're interested in privacy, to attend, look at these sessions, learn about the approach, learn about how we're thinking about creating Chainlink's privacy. I think it will be very, very important for our industry. Privacy is truly one of the last major important steps to getting our industry operating everything at global mainstream scale.

Our goal is truly to ensure that all of these services, data services, messaging services, privacy, identity, all these different service combinations and the coordination of services into advanced contracts is what provides the reliability needed to hold value, is easy for developers to use so they can quickly build high-quality advanced smart contracts efficiently and bring them to market and succeed.

And meet the requirements of regulators, society, and the world at large, so they can achieve widespread global adoption. Generally, Chainlink is seen as giving people choices, the choice about which data to use, which encryption to use, which privacy system to use. We still believe in that.

We want to make CRE and CRE's privacy capabilities and data capabilities as flexible as possible. If you have a specific need from CRE or from Chainlink, please share it with us. We're definitely here to help you.

Developers and those building the next generation of our industry, we're very much looking forward to hearing what you need from us and from the Chainlink community here to help you build the next generation of smart contracts. Things that will take our industry to the next step.

There are many sessions about all of these topics. Too many for me to talk about each one. But we have sessions on how to use AI, cross-chain, regulatory compliance, what we're doing about data for institutions, and generally many sessions, including some on confidential compute. I encourage you to look at these from a product perspective and see what's really being built.

Thank you for your attention and interest in what we've built together here. Thank you for attending this event. I encourage you, in the spirit of Chainlink, to be straightforward. If you meet someone from a financial institution, be nice to them. And if you're from a financial institution meeting a DeFi person, they make a ton of money. They make a lot more money than it seems. So just keep that in mind.

It's been great talking to all of you. Thank you for your interest in our work, and I hope you enjoy the conference.