Blog

Dear imToken users, partners, and friends,Ten years ago, when we started building imToken, there was no clear map of the industry. Blockchain was still in its early days, and crypto wallets had not yet become a mature product category. For most people, private keys, seed phrases, signatures, gas fees, and smart contracts were unfamiliar and difficult concepts. The world did not yet fully understand crypto assets, and few believed that individuals could truly own and control their own value.But we made a choice that later proved to be very important: we chose to build a wallet. We believed that the internet should not only give you accounts; it should also give you assets, identity, and choice. What truly belongs to you should be in your own hands. This is where imToken began.Over the past ten years, we have gone through bull and bear cycles, the rise of DeFi and multi-chain ecosystems, and repeated tests of trust, risk, and long-term thinking together with tens of millions of users. Many things have changed, but one belief has remained the same: imToken has always believed that what truly belongs to you should be under your control.Digital Assets, Under Your Control.This idea sounds simple, but it has been our most important product principle over the past decade. It is also the starting point for where we go next.The First Decade: Store, Send, StakeLooking back on imToken’s first ten years, the core value that wallets created for users can be summarized in three words: Store, Send, and Stake.Store means enabling users to truly hold their own digital assets. Wallets first solved the challenge of self-custody: how to hold assets securely, how to avoid relying on centralized institutions, and how to return final control over value to individuals. Self-custody is not merely a feature; it defines where control resides. It means that assets that truly belong to you do not need to be entrusted to another centralized account.Send means enabling value to move freely. Once users can own assets, the next step is to transfer value freely. Send is not just about making transfers. It is the ability for value to flow across an open network, allowing anyone to move what belongs to them across geographic and institutional boundaries without needing permission.Stake means helping users move from holders to participants. As open networks evolved, wallets became more than tools for storing and transferring assets. They became gateways for users to participate in networks, earn rewards, and support ecosystem operations. From staking to DeFi, from governance to broader forms of on-chain participation, wallets helped users move from “owning assets” to “participating in networks.”Store, Send, and Stake have formed the core product value of imToken’s first decade. Over these ten years, imToken has grown much like a tree. Our mission, vision, values, and long-term principles are the roots of this tree; our products, technology, security, user experience, and services are the branches growing upward. Only deep roots can support flourishing leaves. And the farther we go, the more clearly we understand what must never be lost.The tenth anniversary is not an endpoint. It is a new beginning after looking back. As we stand here today and look toward the next decade, a larger shift is taking place.The Intelligent Age: Abundant Agency, Scarce ControlAI is bringing the internet into a new era. In the future, more and more software will no longer simply wait for people to click buttons. It will understand goals, break down tasks, call tools, and act on behalf of people. We are entering the age of intelligent agents: AI agents will have increasingly powerful agency, enabling them to search, reason, trade, collaborate, pay, and execute in the digital world.This will bring enormous productivity, but it will also create new risks. In the past, the core question of the digital world was: How can people own assets? In the future, an increasingly important question will be: When intelligent agents can act on behalf of people, how can people still remain in control?We see three forces converging. First, AI is sweeping across the world, transforming software from “waiting for human operation” to “acting on behalf of humans.” Second, the value of blockchain is becoming visible again. Permissionlessness, censorship resistance, self-custody, verifiability, and traceability were once foundations of open finance. In an AI-native internet, they will also become important defensive infrastructure for collaboration between humans and intelligent systems. Third, control is becoming a new personal challenge. In the future, every individual will not only manage assets, identity, and data, but also their own AI agents, automated tasks, and authorization boundaries.Agency will become increasingly abundant. What will truly be scarce is control.This is not an abstract issue. In the future, users will need to know: Who can act on my behalf? What are they allowed to do? What is the spending limit? How long does the authorization last? Which actions require my confirmation? When something abnormal happens, how can I pause, revoke, and trace it? Without clear mechanisms of control, intelligent agents may become a new kind of black box. The more powerful they become, the easier it will be for people to lose boundaries, accountability, and final decision-making power.So for imToken, the mission of the next decade is not to build another AI agent, nor is it simply to add AI features to a wallet. What we truly care about is this: In an AI-native internet, how can people still retain final control?The Fourth S: SignIf the 3S of the first decade were Store, Send, and Stake, then the fourth S for the next decade is Sign.In the past, Sign usually meant signing a transaction. But in the age of intelligent agents, the meaning of Sign will expand. It will no longer be just a transaction approval. It will become a foundational interface through which humans express authorization, define boundaries, delegate actions, and retain control.In the future, what you sign may not only be a transaction. It may be an intent, a permission and delegation, a policy, or a revocation.AI agents can propose plans, execute tasks, and amplify efficiency. But their actions must be built on top of your clear, verifiable, constrained and revocable authorization.Sign to permit intelligent actions under your control.This is how we understand the fourth S. It is also the core product thesis for imToken’s next decade.From a Trusted Main Wallet to a Personal Control InterfaceOver the past ten years, imToken has become a trusted main wallet for many users. We treat this with deep respect. “Main wallet” is not a title we take lightly. It means users have entrusted us with long-term assets, important identities, critical transactions, and a sense of security. It also requires us to remain restrained, reliable, and long-term oriented in product, technology, security, user experience, and risk communication.In the future, imToken will still first and foremost be a trustworthy wallet. We will not abandon the clarity of what a wallet is in order to chase concepts, nor will we push users into complex systems they cannot understand or control. But we also know that the role of the wallet is naturally expanding. In an open and intelligent internet, individuals need to manage not only assets, but also identity, permissions, data, AI agents, automated tasks, and increasingly complex digital relationships.Therefore, imToken will gradually evolve from a trusted main wallet into a personal control interface — an interface that helps you control your assets, identity, permissions, and intelligent actions.This is not a leap from wallet to abstract platform. It is a natural extension of the wallet’s core logic: keys are the root of control; signatures are the expression of authorization; permissions define boundaries; policies set the rules for actions; revocation protects freedom; and auditability is the foundation of accountability.The wallet of the future will not only be an entry point for assets. It will also become a trusted control interface for collaboration between humans and intelligent systems.The Direction of the Next DecadeGuided by this direction, imToken will continue to protect the root of self-custody. No matter how technology changes, users should not easily give away final control over their assets and critical rights.We will also move from transaction security toward authorization security and action security. In the past, users were most concerned about private key leaks, signing the wrong transaction, or interacting with malicious contracts. In the future, users will also need to know whether they have authorized an agent to act over the long term, whether they have granted excessive limits, and whether they can pause or revoke that authorization at any time. What imToken needs to do is make these originally complex, hidden, and hard-to-understand authorization relationships clearer and more controllable.AI can help people accomplish more. But people should always be able to understand, approve, limit, and audit those actions. We want to make intelligent actions more controllable, not make control more ambiguous.This is a long-term direction, and it is not a question that imToken can answer alone. But it is a question we must help build toward.Building for the Long Road AheadTen years is not an endpoint. It is a clearer starting point.We will continue moving forward as builders: fewer slogans, more products; less chasing of hype, more infrastructure; less short-term noise, more long-term trust. We will also continue to stay true to why we started, sense the direction of change, understand what is essential, endure through time, keep iterating, and move forward with conviction.If imToken’s first decade was about helping you truly own your digital assets, then imToken’s next decade is about helping you continue to control your digital world in the intelligent age.Your Digital World, Under Your Control.Thank you to every user, developer, investor, partner, and team member who has experienced, believed, and built with imToken over the past ten years.The road ahead is still long. We will continue exploring, continue building, and continue moving toward the horizon. Ben HeFounder & CEO, imTokenJune 2026

For a long time, when we talked about wallets, we were mostly talking about assets.Where should BTC be stored? How do you transfer ETH? How should NFTs be managed? How do users access and use DeFi or RWAs? For most crypto users, a wallet was, in a sense, their gateway to assets.But AI is changing this.When users can describe what they need in natural language, and when AI can help break down the steps required to complete an action, the role of the wallet also begins to change. This has become especially clear over the past six months. A wallet is increasingly becoming the command center for a user’s digital world.From this perspective, the real question for wallets in the AI era may not be whether they can do more things on behalf of users. It is this: when more and more actions can be automated, how can users continue to understand each interaction and retain ultimate control?This is the new question imToken continues to answer as it enters its next decade.1. The New Wallet Narrative: From an Asset Entry Point to a Personal Digital HubIf you had told an Ethereum user in 2016 that ten years later, they could simply type into a chat box, “Help me generate a minimalist wallet that only shows NFTs, AI-related tokens, and common actions,” and then receive an app that could run on a testnet, they would probably think you were a project founder who could not even write a convincing whitepaper.But by 2026, this no longer feels like science fiction.If you recently took part in imToken’s 10th anniversary event, you may have seen something similar already becoming possible: a user only needs to describe their need in natural language, and an initial wallet interface can be generated, showing NFTs, AI tokens, and common actions such as Receive, Sign, and Swap.“Your Digital World, Under Your Control” is a fitting way to summarize imToken’s new narrative for its tenth anniversary. It is not about packaging the wallet as a platform that does everything. Rather, it recognizes that as the digital world users enter becomes more complex, they need a long-term, trusted, secure, and clear entry point that remains under their own control.That entry point has been the wallet, and the wallet will continue to evolve into it. The more complex the digital world becomes, the more it needs a trusted starting point.In the past, wallets mainly helped users prove that “these assets belong to me.” Whether it was ETH, ERC-20 tokens, NFTs, DeFi positions, or later RWA assets, the wallet’s core role was to serve as an asset container and a signing entry point.But in the AI era, wallets also need to help users confirm more things: Do these identities belong to me? Are these authorizations managed by me? Do I understand these actions? Are these automated workflows still within my control boundaries?This is the core of the “personal digital hub” narrative. It also means the wallet’s next stage is not merely to remain a wallet, but to become the foundational interface for entering the digital world.Take imToken as an example. If we divide its past decade into three stages, a clear trajectory emerges:From 2016 to 2023, the wallet was a container for assets. Starting from the Ethereum ecosystem and expanding alongside ERC-20 tokens, DeFi, NFTs, and other asset forms, the core question was simple: how to keep private keys as safely as possible on users’ own devices, and how to make every newly emerging token reliably fit into the same container. At this stage, users cared most about whether assets could be stored safely and accessed or moved out smoothly.From 2024 to 2025, wallets began to stand at a paradigm shift. Tokens were no longer just assets. They started to extend toward identity, data, agents, and permission relationships. Ethereum’s narrative also moved beyond scaling and toward directions closer to user experience, such as account abstraction. As the way users interact with blockchains began to be rewritten, the wallet, once a relatively stable piece of the puzzle, began to shift significantly for the first time.From 2026 onward, wallets are moving toward the role of a “personal digital hub.” As AI begins to participate in app generation, transaction understanding, risk detection, and automated execution, the wallet is no longer just a tool being used. It is becoming more like each person’s digital command center, coordinating collaboration between users and AI agents.These three stages can be summed up in one sentence: tokens evolve, control remains.Asset forms will change. Interaction models will change. AI capabilities will change. But what the wallet must protect remains the same: the user’s ultimate control over their own digital world.2. Functionality Is Not the Destination. Security Is the Foundation.Take imToken’s 10th anniversary AI co-creation initiative as an example. What matters most is not simply “generating a wallet interface with AI,” but how it brings the question of how wallets should work with AI down to a more foundational level.One thing needs to be made clear first: the AI direction imToken has shown so far is not the radical path of “handing private keys to AI and letting AI trade automatically for you.” Instead, it focuses on three more practical directions: allowing users to participate in wallet co-creation through natural language, making the wallet’s underlying capabilities easier for developers and AI to access, and embedding security rules into generation and interaction from the start.We believe this path is more consistent with how wallets should evolve.Because a wallet is not an ordinary app. If an ordinary app gets a button wrong, the result may simply be a poor user experience. But if a wallet gets a signature, an authorization, or a private key handling process wrong, it may lead to real asset loss. That is why wallets in the AI era cannot focus only on generating things quickly. They must also be secure by design, easy to understand, and verifiable.One of the most concrete steps is to further open Token Core capabilities to co-creation scenarios. For everyday users, Token Core may sound technical, but it can be understood as the “heart” of the imToken wallet. It handles the wallet’s most essential capabilities, such as private key and keystore management, address generation, transaction signing, and multi-chain support.In simple terms, wallet interfaces can take many forms. But what truly determines whether a wallet can securely manage assets, sign correctly, and run reliably across different chains is this underlying “heart.”Token Core was open-sourced as early as 2018. At the time, it mainly served imToken’s own mobile wallet, supporting multi-chain asset management and signing on iOS and Android. Today, Token Core has evolved into a wallet core library that supports multiple public chains and cross-platform calls.What is especially worth noting in the tenth anniversary-related branch is the emergence of a WebAssembly version.WebAssembly sounds technical. In plain language, WebAssembly makes it easier for core wallet capabilities that used to run mostly in apps or local environments to run in the browser. As a result, web-based wallet demos, AI-generated wallet applications, and wallet prototypes built by developers may be able to call underlying wallet capabilities more directly.The significance is that the wallet no longer has to be just a collection of features inside a closed app. It can become a more open and composable set of foundational capabilities. Alongside this, several easier-to-understand tools have also appeared: The Token Core CLI demo can be understood as a “command-line demo console.” It breaks down core wallet actions, such as creating a wallet, deriving addresses, managing keystores, and signing transactions, so developers and AI can see more clearly what the wallet is doing under the hood. Token UI can be understood as a “wallet interface template library.” Built on imToken’s design system, it helps participants build wallet-like interfaces more quickly. Users can ask AI to generate a wallet interface prototype without having to design every button, list, and asset card from scratch. security/SKILL.md is more like a “wallet security manual” written specifically for AI coding assistants. When AI generates code involving seed phrases, private keys, signatures, or authorizations, it cannot simply focus on making the feature work. It must first understand the hard boundaries, and any asset-related operation must require user confirmation. These open-source initiatives may differ from how many people used to think about wallet competition.In the past, it was easy to think of a wallet as an app: whoever supported more chains, had a better-looking interface, and offered more complete DApp entry points had the advantage. But after the arrival of the AI era, wallet competition may take a different form: whoever can provide more trusted underlying capabilities, help users and developers compose wallet functions more safely, and still maintain security boundaries when AI generates experiences will be better positioned to become the foundation of users’ digital worlds.This is why imToken’s AI efforts should not be reduced to “an AI wallet-generation event.” It is actually answering a more foundational question: when AI can generate more wallet interfaces, interactions, and applications, what must remain stable? What can be opened to users and the community for recomposition? And what must be constrained by security rules?imToken’s answer is: trust belongs in the core, control belongs to the user, and innovation belongs to the community.3. The New Map for Crypto Users: From Natural Language Entry Points to Agent Boundary ManagementWhat, then, can we expect Web3 wallets to look like over the next decade?If we bring the two threads above together, one side is imToken bringing the wallet core, UI templates, and security rules to users and developers. The other side is AI gaining stronger understanding and orchestration capabilities between users and blockchains. In this process, the position of an ordinary crypto user is undergoing a very interesting shift.In the past, users mostly adapted to wallets.Users used whatever homepage the wallet provided, clicked the features it supported, and followed its transaction flow step by step. Even heavy users were often just switching back and forth between fixed functions.But with AI involved, wallets may increasingly adapt to users. This means the Web3 wallet of the next decade may not simply have more and more features. Instead, its form may become increasingly personalized. You may no longer need to tolerate a wallet homepage that looks the same for everyone. If you are a DeFi power user, you could ask AI to generate a minimalist interface focused only on yield, risk, and position changes, bringing together major positions across chains, yield rates, redemption timelines, and risk status. If you only care about stablecoin inflows and outflows, your wallet homepage could show only your USDC and USDT balances, recent incoming payments, and frequently used receiving addresses, without distracting you with irrelevant assets and entry points. If you are deeply involved in LSTs or LRTs, the wallet could bring the real ETH positions, yields, exit windows, and potential risks behind different staking tokens into a clearer dashboard. If you simply want to set up a small wallet for a family member, it could keep only receiving, sending, and balance display, while hiding complex DApps, authorizations, and cross-chain features. The underlying signing, address, and transfer logic does not change. What changes is the upper-layer experience. In short, a wallet is no longer just a standardized product. It becomes a digital tool assembled from the wallet core, UI Kit, and personal needs.Looking further ahead, the next generation of crypto users may enter an on-chain world populated by many AI agents.Your AI assistant may scan stablecoin pool spreads for you every day. Your research agent may run small tests when a new protocol launches. Your payment agent may handle subscriptions, refunds, and payment splitting. Your asset management agent may remind you to rebalance according to rules you set.These scenarios may sound futuristic, but they do not mean users should hand their private keys to AI. Quite the opposite: the stronger agents become, the more important wallets become. A healthy relationship between AI and wallets is not about letting agents take unlimited control of user assets. It is about allowing agents to make requests, while the wallet translates those requests into transaction details users can understand and hands the final confirmation back to the user.In other words, AI agents can discover opportunities, make suggestions, and generate paths. The wallet must be responsible for risk warnings, permission constraints, and final signatures.Overall, AI will make wallets smarter and on-chain actions smoother. This is a major shift, and it has only just begun.Final ThoughtsThe underlying logic of the crypto world has always been built on user control. The question of private keys will not disappear because of AI. On the contrary, it will become even more important.This is where imToken’s new narrative, and the truly important direction of the wallet sector, come into focus.Especially as the digital world expands from assets to identity and AI agents, users will still need a trusted entry point that helps them understand, confirm, and control every digital action they take. From a trusted main wallet to a personal digital hub, this is not conceptual packaging. It is the natural extension of the wallet’s role in a new technological environment.Perhaps when we look back at 2026 from 2036, we will see a somewhat counterintuitive fact: the next decade of wallets will not only be about more powerful features. Users will no longer simply be the people services are built for; they will become the ones who define those services.Your digital world, under your control.

For a long time, when we talked about wallet security, users were most often reminded of two things: keep your mnemonic phrase safe, and do not click phishing links.This is because in a self-custodial wallet, the mnemonic phrase or private key always represents control over assets. Its importance cannot be overstated. However, as AI Agents begin to enter wallets, trading, payments, and on-chain execution, a new issue is becoming increasingly important: even if your private key is never exposed, your assets may still be transferred because of an incorrect approval, a misleading signature, or a compromised automated instruction.In other words, wallet security is moving beyond “who controls the assets” toward more specific questions: why are the assets moving, how are they being moved, and does the action truly match the user’s intent?This is also why Clear Signing was further advanced into Ethereum’s open standardization process on May 12. To be clear, it is not trying to solve a new problem, but a long-standing one in crypto: many users are not careless about security. They simply cannot understand what they are signing before clicking “Confirm.”1. In the AI Agent Era, Web3’s Security Boundary Is Quietly StretchingWith the rise of AI Agents, Web3 on-chain interactions are also moving closer to natural-language experiences.In the past, if you wanted to complete an on-chain action, you had to open a DApp, connect your wallet, choose a route, approve permissions, initiate a transaction, and confirm every step yourself through wallet pop-ups. In the future, this process may be greatly simplified into a single sentence: “Help me find a stablecoin pool with higher yield,” or “Claim this airdrop and swap it into ETH.”From a user experience perspective, this is clearly progress. AI Agents can help users understand information, break down steps, generate transactions, improve efficiency, and even complete certain actions automatically within defined permissions.But the other side of improved efficiency is an expanded security boundary.At that point, what determines where funds go is no longer just the user. It may also involve the Agent’s interpretation, external data sources, and several other steps in the execution chain. If any one of these steps is compromised, what the user sees as “help me execute this” may become what the attacker wants: “transfer funds on my behalf.”Recently, attackers used prompt injection on X to induce AI Agent systems to execute suspicious transfers involving 3 billion DRB tokens, valued at roughly $150,000 to $200,000. The core issue in incidents like this is not traditional private key leakage, but how AI systems interpret inputs, how they obtain permissions, and how instructions are passed to the on-chain execution layer.This also shows that attackers do not necessarily need to break into a wallet directly. If they can make an Agent operating with excessive permissions mistake malicious input for a valid command, real financial losses may follow.After all, in traditional internet scenarios, prompt injection may only cause an AI system to give a wrong answer, leak context, or call the wrong API. But in crypto, once an Agent is connected to a wallet, has permissions, and can initiate transactions, a wrong instruction may directly become an on-chain transfer. Since on-chain transactions are irreversible, AI Agent security is no longer just about “model security”; it is asset security.Therefore, wallet security in the AI Agent era cannot rely only on making AI “a little smarter.” The real key is that between the transaction generated by the Agent and the signature confirmed by the user, there must be a security interface that is clear, verifiable, and understandable.That interface is the wallet.2. Does Clicking “Confirm” Really Mean the User Understands?For most users, the most familiar wallet action is probably “Confirm.”Connecting to a DApp requires confirmation. Swapping requires confirmation. Token approvals require confirmation. Bridging requires confirmation. Claiming an airdrop requires confirmation. Staking, lending, or minting an NFT also requires confirmation.The problem is that many confirmation pages do not truly tell users what will happen after they confirm.In many cases, users only see a function name. Sometimes they see a block of unreadable hexadecimal data. Sometimes they only see a vague “Approve” or “Sign Message.” Technically, this information may not be wrong. But for most users, it is not enough to make an informed decision.This is the danger of blind signing.Blind signing does not mean the user did not look at anything at all. It means the information shown is not enough for the user to make an informed decision. It is like signing a contract written in a language you do not understand, with only an “Agree” button visible at the end. You know you are signing, but you do not know what consequences the signature will bring.In its Clear Signing announcement, the Ethereum Foundation also emphasized that the final step in many major attacks is not a code vulnerability, but a user approving a transaction they cannot truly understand. If transaction confirmation is supposed to be the final line of defense for user control over assets, then blind signing makes that line of defense ineffective.So if account abstraction over the past few years has focused on “how to execute more conveniently,” then Clear Signing focuses on “how to verify more clearly before execution.” These two are two sides of the same coin. Without better signature interpretation, more complex automation and more powerful account capabilities may also create more room for user error.This is where ERC-7730 comes in. According to the EIP-7730 proposal, it defines a structured data format for Clear Signing. By using JSON files to supplement information beyond ABI and message types, it transforms raw transaction data into content that is easier for humans to verify, while also allowing machine systems such as transaction simulation tools to use it directly.Put more simply, ERC-7730 does not change the on-chain transaction itself. Instead, it adds a standardized explanatory layer between the transaction and the user. For example, as shown below, a wallet may previously have only been able to display function selectors and parameters. With ERC-7730, it can present the action in a way users can actually understand.With this standard, any wallet that supports ERC-7730 can display raw function selectors and numeric parameters as human-readable content such as “Swap 1,000 USDC for at least 0.42 WETH.” This may look like a UI improvement, but in reality, it is a fundamental upgrade to wallet security.Only when users can understand the transaction does confirmation have real meaning. And only when wallets can present transaction intent in a structured way do users have a chance to identify problems before signing.3. Verifiable UI: Making What Users See Match What Will Actually HappenThis brings us back to the Verifiable UI concept we have been emphasizing recently.If Clear Signing aims to help users understand what they are signing, Verifiable UI goes one step further: can the content users see be reliably mapped to actual on-chain execution?This is critical in Web3.Many users are used to trusting DApp frontends. If the page says “claim rewards,” they assume they are claiming rewards. If the page says “stake,” they assume they are staking. If the page says “security verification,” they assume it is only identity verification.But what can actually move assets is not the button on the webpage. It is the transaction ultimately signed inside the wallet.A DApp frontend may be attacked. A domain may be spoofed. Page copy may be disguised. Even the information read by an AI Agent may come from a compromised webpage or social post. If the wallet simply shows a generic confirmation prompt, users are still in a state of “trusting the frontend.”This is also why imToken’s plan to support ERC-7730 and advance Verifiable UI + Clear Signing is important.It is not simply about showing a few more lines of text on the confirmation page. It is about turning the wallet from “the last click in a transaction flow” into “the final layer of verification before signing.” When a user or AI Agent is about to initiate a transaction, the wallet should tell the user as clearly as possible which contract the transaction actually calls, which asset is being transferred, who the approval is granted to, how broad the approval scope is, and whether the final result matches what the page displays.This capability will become even more important in the AI Agent era.Agents can help users do many things, but Agents can also make mistakes. Users cannot hand all judgment over to Agents, and wallets should not simply pass Agent-generated transactions to users for confirmation without interpretation. A more reasonable division of responsibilities is this: Agents improve efficiency, while wallets guard the boundary.This is the value of Verifiable UI + Clear Signing. It is not meant to prevent users from using new technologies. It is meant to let new technologies operate within more verifiable boundaries. As smart accounts, AI Agents, automated trading, and cross-chain execution become more common, wallet confirmation pages should no longer remain in a low-information state of “Confirm / Approve.” They should become a key interface through which users understand on-chain actions.Further reading: “From Kelp DAO to Verifiable UI: The Next Security Baseline for Decentralization”Final ThoughtsThe crypto industry has always pursued a better user experience.From mnemonic phrases to smart accounts, from manual operations to AI Agents, and from single transactions to batch execution, wallets are becoming more powerful and more similar to everyday internet products in how they are used. But the more this happens, the more we must not ignore one basic fact: on-chain transactions are irreversible, and signatures remain the most critical step before user assets move.In the past, we often said: “Do not leak your mnemonic phrase.” In the future, as AI Agent capabilities become widely embedded in Web3 and on-chain execution, we may need to add another reminder: do not sign transactions you cannot understand, and do not let Agents execute instructions you cannot verify.Ultimately, whether it is the Ethereum Foundation promoting the standardization of Clear Signing, or imToken planning to support ERC-7730 and advance Verifiable UI + Clear Signing, both point in the same direction:In the new era, wallets should not only be easier to use. They should also be more trustworthy — becoming a reliable guide that helps users understand what is happening on-chain.

For many users, their most direct impression of Ethereum in recent years has not come from roadmaps or developer conferences, but from the on-chain actions they perform again and again.For example, many users have felt transfer gas fees becoming lower over the past two years, and cross-chain interoperability becoming smoother. This is why Ethereum scaling has never been just a “performance race.” For everyday users, higher TPS, larger blocks, and more complex architecture only matter when they translate into lower costs, smoother interactions, and a safer wallet experience.Recently, a series of Ethereum developments have pointed in the same direction: Ethereum is trying to systematically shift some of the complexity previously handled by wallets, DApps, third-party relayers, and users themselves down to the protocol layer.These include Keyed Nonces, a proposal Vitalik has been involved in; the directional consensus around a 200 million gas limit floor in the Glamsterdam upgrade; and a set of connected signals in the 2026 roadmap, including native account abstraction, cross-L2 interoperability, and stronger L1 security.1. Raising the Gas Limit to 200 Million?Let’s start with the part users can notice most directly: the gas limit.Every Ethereum transaction — whether a transfer or a contract interaction — consumes gas. Each Ethereum block has a fixed gas limit, meaning block space is limited: the more space there is, the more “passengers” can be carried at once; when space is tight, users bid for the same seats, and gas fees naturally rise.In theory, increasing the block gas limit can significantly improve Ethereum mainnet performance. But against the backdrop of rapid L2 development, Ethereum has remained cautious, intentionally channeling much of the scaling pressure toward L2s.Looking at Ethereum’s gas limit history, after the network first rose from 8 million to over 10 million in September 2019, it took roughly seven years to reach 60 million. The real acceleration came in 2025: from 30 million to 36 million in February, then to 45 million in July, and finally to 60 million after the Fusaka upgrade in December.In other words, most of this expansion was concentrated in 2025. As we noted before, 2025 was also a critical year in Ethereum’s history. The Fusaka upgrade, coming just seven months after Pectra in May, showed that even after major leadership changes at the Ethereum Foundation, Ethereum could still deliver major upgrades. It also marked Ethereum’s move into an accelerated cadence of “two hard forks per year.”Related reading: “Ethereum 2026: Interpreting EF’s Latest Protocol Roadmap—Is Ethereum Entering an Engineering-Driven Upgrade Era”Source: EtherscanAccording to the Ethereum Foundation’s Soldøgn Interop Recap published on May 2, more than 100 Ethereum core contributors gathered in Svalbard, Norway, for an interoperability meeting focused on Glamsterdam. The meeting aimed to advance multi-client implementation, testing, and parameter alignment. By the end, developers had reached broad consensus around a 200 million gas limit after Glamsterdam.If the process goes smoothly, Ethereum L1 execution capacity could rise from today’s roughly 60 million gas limit to around 200 million. Over a longer horizon, the Ethereum ecosystem has clearly become more open — even more aggressive — in publicly discussing gas limit increases. EIP-9698 even proposes a 10x increase every two years, raising the gas limit to 3.6 billion by 2029, about 50 times today’s level.But it is important to emphasize that increasing the gas limit is not simply about making blocks bigger.If each block’s computational capacity is increased in a brute-force way, fees may fall in the short term. But over the long term, it could increase node burdens, accelerate state growth, make it harder for ordinary users to run nodes, and ultimately weaken Ethereum’s core foundation of decentralization.That is why Glamsterdam’s scaling approach is a coordinated package: ePBS, or enshrined Proposer-Builder Separation, brings block building and validation more clearly into protocol rules, allowing validators to handle larger blocks more safely. Block-Level Access Lists, or BAL, record in advance which accounts and storage locations will be accessed during block execution, enabling parallel disk reads, parallel transaction verification, and parallel state root computation. EIP-8037 increases the cost of state-creating operations, helping prevent state growth from accelerating too quickly after the gas limit is raised. In the end, Ethereum is not only trying to include more transactions. It is also asking how to do so without making node operation increasingly difficult.This is the fundamental difference between Ethereum’s scaling roadmap and many high-performance-chain narratives. Ethereum is not trying to trade higher verification costs for superficial throughput. Instead, it aims to increase mainnet capacity while preserving ordinary node participation and system verifiability as much as possible.2. Keyed Nonces: Turning “One Queue” into “Multiple Channels”If the gas limit answers “how much can fit into one block,” then Keyed Nonces focus on a more detailed but equally important question: how should a transaction wait in line?In Ethereum, a nonce can be understood as the “sequence number” of an account’s transactions. It prevents the same transaction from being executed twice and ensures that transactions from the same account are processed in order.This mechanism is easy to understand in simple transfer scenarios: the first transaction, the second transaction, the third transaction, and so on, all lined up in order.But as account capabilities become more complex — involving privacy transactions, smart wallets, session keys, batched operations, and third-party gas sponsorship — a single linear nonce can become a bottleneck. EIP-8250’s Keyed Nonces address this by allowing an account to have multiple nonce domains instead of only one nonce queue.More specifically, it replaces the single sender nonce in EIP-8141 Frame Transactions with a (nonce_key, nonce_seq) structure. Here, nonce_key == 0 corresponds to the traditional account nonce, while non-zero keys can use independently managed nonce sequences at the protocol level. Transactions under different keys are independent, and their replay protection does not interfere with one another.This may sound technical, but a simple analogy helps: in the past, an account was like having only one service counter at a bank, where every type of task had to wait in the same line. Keyed Nonces are like opening separate counters for different tasks, so transfers, private withdrawals, session authorizations, and batch executions can each use their own channel.This is especially important for privacy protocols. To avoid directly linking users’ on-chain activity to a single public address, a privacy protocol may allow multiple users to send transactions through the same shared sender address. With a single nonce, once one user’s transaction is included, other users’ pending transactions may become invalid or blocked.Keyed Nonces allow each transaction or spend to choose its own nonce domain — for example, one derived from a privacy nullifier — reducing this kind of queueing conflict at the protocol layer.Vitalik frames it even more broadly. When introducing EIP-8250, he said Keyed Nonces are “not only stronger support for protocol-layer privacy schemes, but may also be the first step toward a new state-scaling strategy for Ethereum: creating storage types optimized for different use cases, enabling very high scalability while preserving protocol decentralization.”Put simply, the gas limit addresses the “size of the block,” while Keyed Nonces explore the “shape of state.” What Ethereum needs to support in the future is not just more transactions, but more types of transactions.3. How Will This Affect Everyday Users?For the Ethereum ecosystem, many protocol upgrades may seem far removed from everyday users. But in the end, they all land in the wallet experience.Users do not experience Ethereum through EIPs, clients, or developer meetings. They experience it through every transfer, approval, signature, cross-chain action, and DApp interaction inside their wallet.In other words, protocol-layer changes only become true user experience upgrades when wallets translate them into clearer, smoother, and safer interactions.Account abstraction, now widely discussed, is a good example. Its purpose is not to make users learn more technical terms, but to help them use on-chain accounts more naturally. That is why batch transactions, gas sponsorship, recovery mechanisms, different signing methods, session authorization, and more flexible security policies are gradually becoming basic wallet capabilities.Keyed Nonces are similar. They sound like a low-level optimization to an account’s transaction queueing mechanism, but from the user’s perspective, their potential impact is not abstract at all.Many users have likely encountered this: one transaction remains unconfirmed for a long time, later transactions get stuck, and the user wants to cancel or speed it up but does not understand the relationship between nonce, gas, and replacement transactions. When multiple actions are happening in parallel, one failed step can affect everything that follows.To everyday users, these issues may look like “the wallet is hard to use” or “the chain is hard to use.” But behind them is Ethereum’s account model, where a single linear nonce determines execution order.Keyed Nonces point toward a model where accounts no longer have to execute every action through a single queue. Instead, they can split different use cases into multiple parallel channels.In theory, regular transfers, DApp approvals, privacy transactions, batch transactions, gas sponsorship, and other actions could each have more independent execution space, reducing the chance that they block or conflict with one another.This will undoubtedly open up more design space for smart wallets.More importantly, these capabilities used to require wallets, DApps, relayer services, and users to share the complexity. Users had to understand authorization scope, judge whether gas was reasonable, know exactly what they were signing, and repeatedly confirm multi-step actions such as bridging, swapping, staking, and claiming rewards. Any misunderstanding along the way could lead to failed operations or asset-loss risk.What Ethereum is trying to do now is shift part of this complexity into the protocol layer, so wallets can build better interaction abstractions on top of more standardized, native capabilities.That is why gas limit, BAL, ePBS, Keyed Nonces, Frame Transactions, native account abstraction, and cross-L2 interoperability may seem like separate technical modules, but they are all serving the same goal: enabling Ethereum to support more complex on-chain use cases without sacrificing decentralization or security.Looking at these developments together, Ethereum’s recent priorities are more coherent than they may first appear: Gas limit increases address mainnet execution capacity and fee pressure. BAL, ePBS, and EIP-8037 address how to preserve node verifiability and keep state growth under control during scaling. Keyed Nonces and Frame Transactions address protocol-layer bottlenecks in the account model, privacy protocols, and smart wallets. Native account abstraction and cross-L2 interoperability point toward experience improvements that everyday users can actually feel. This also means Ethereum is entering a new stage.Over the past few years, the market has focused more on L2 scaling, blob-driven fee reductions, and the modular blockchain narrative. Users have also grown used to moving assets across L2s and looking for lower-cost environments to interact on-chain.But as the mainnet gas limit continues to rise, upgrades such as Glamsterdam move forward, and account abstraction and interoperability solutions continue to evolve, Ethereum is no longer only asking “how to make transactions cheaper.”It is also asking: “How can the on-chain experience feel more like one coherent whole?”Along the way, wallets will become even more important.Wallets are not only the entry point into Ethereum; they are also the interface through which users understand and use protocol capabilities. The more complex the underlying upgrades become, the more wallets need to translate them into clearer signing prompts, more understandable transaction paths, earlier risk detection, and smoother on-chain interactions.Let’s keep building toward that future.

Ten years is a moment for reflection. More importantly, it is a moment to reaffirm where we are heading.In the past, the digital world in crypto was largely centered around assets. From BTC and ETH to ERC-20 tokens, NFTs, DeFi, and RWAs, the forms of token have continued to expand, and wallets have helped users securely hold, manage, and use these assets.Now, AI is reshaping digital infrastructure and the way people interact with it. As users’ relationship with the digital world become more flexible, efficient, and intent-driven, they are also becoming more complex. “You” are no longer just an address or an account, but a more complete digital presence — carrying identity, permissions, ownership, and boundaries of action into the open network. And the “digital world” is no longer limited to assets. It also includes identity, authorization, smart accounts, AI agents, and more automated interaction scenarios.That is why imToken needs to continue evolving. What remains unchanged is “under your control.” No matter how far the digital world expands, users should retain ultimate control over their assets, identity, and actions.imToken’s brand evolution starts from its role as a trusted main wallet. Built on Trust, Control, and Becoming, imToken is gradually evolving into a personal control interface for the open, intelligent internet.A Wallet Begins with TrustFor many users, a wallet is not just another app.An ordinary app can be replaced, uninstalled, or re-registered. A wallet is different. It holds the digital assets users truly own. It is the starting point of every on-chain action, and the entry point through which users build their relationship with the open network.That is why the core value of a wallet is not how many features it has, but whether it can be trusted.Over the past ten years, imToken has built around this foundational value: users hold their own private keys and control their own assets; assets are not held by the platform, nor is ownership defined by a centralized account. A wallet needs to be reliable, secure, stable, and worthy of being trusted with users’ primary assets.This is imToken’s most important product role today: a trusted main wallet for your digital world.This positioning is simple, but important. In an open network, true control is never just a slogan. It is supported by a series of concrete capabilities: users need to manage private keys securely, confirm signatures clearly, understand what they are authorizing, know what each transaction may result in, and receive clear risk signals in a complex on-chain environment.These specific and foundational capabilities form the basis of imToken’s trust. Ten years later, we still believe that wallet is imToken’s clearest and most important starting point. Now, we are building on top of it to answer the next question.The User’s Digital World Is Opening Up New PossibilitiesIn the past, when we talked about wallets, we were mostly talking about assets: which tokens users hold, how they participate in DeFi, how they manage NFTs, and how they complete an on-chain transaction.Today, AI, smart accounts, cross-application authorization, and richer on-chain applications are opening up more possibilities for users’ digital worlds. A wallet will no longer only help users store and transfer assets. It will also take part in identity verification, permission management, application access, and more digital actions assisted by AI agents.The digital world is becoming more capable and more automated. Users can enter applications more easily, call services, combine assets and identity, and let systems assist with complex operations that used to require manual effort.When more actions can be authorized, delegated, or executed automatically, users need to understand more clearly: What identity am I using to enter an application? Who have I authorized? Which permissions are still active? Which actions represent me? Who holds the final right to confirm and control?These changes will continue to drive the evolution of the wallet. A wallet will no longer be just an asset container. It will increasingly become the central interface of a user’s digital world: connecting assets and identity, managing transactions and authorization, and helping users stay in control across a broader digital world.From a Trusted Main Wallet to a Personal Control InterfaceThis is the direction imToken is moving toward: from a trusted main wallet to a personal control interface.In the past, that entry point was the wallet. In the future, the wallet will remain the foundation of that entry point.imToken remains committed to its legacy: a trusted main wallet for long-term self-custody, where users retain full control and true ownership of their assets. At the same time, we are building for what’s next. We are transforming imToken from a secure vault into a dynamic interface—supporting not just your assets, but your identity and every intelligent interaction in the open network.In short, your main wallet today. Your personal control interface tomorrow. Trust, Control, and Becoming‘Under your control’ is not just a slogan—it is a reality built on three unbreakable pillars: Trust, Control and Becoming.Trust is the foundation for a wallet to carry users’ assets over the long term. Without trust, users will not entrust a product with important assets, key identities, or high-value permissions.Control is the core that imToken has always upheld. In the past, this control was mostly reflected in asset self-custody. In the future, control will become broader. As identity, permissions, smart accounts, AI agents, and more on-chain behaviors enter users’ digital lives, users need to control not only “where the money is,” but also “who can act on my behalf,” “which permissions belong to me,” and “which actions require my final confirmation.”Becoming is our commitment to evolution. We are building beyond the traditional wallet to create the ultimate personal control interface. It is a gateway designed to solve the friction of the next digital era, turning complex permissions and interactions into a seamless, user-controlled experience.Our journey is defined by pragmatic innovation, and these are our commitments.Your Digital World, Under Your ControlThe tenth anniversary is not an endpoint. It is a new beginning.Over the past ten years, imToken has accompanied users as they entered the on-chain world, managed assets, understood self-custody, and built trust in the open network. In the next ten years, users’ digital worlds will continue to expand. Assets will become more diverse. Identity will become more important. The Permission relationship will become more complex. AI agents will participate in more digital actions. And the open network will carry more real user needs.What imToken wants to uphold remains unchanged: users should own their digital world, understand their digital actions, and control their assets, identity, and permissions.So this brand evolution is not a conceptual shift. It starts from the trusted main wallet and continues to answer the same question:In an ever-changing digital world, how can users remain in control?Your Digital World, Under Your Control. 

The on-chain DeFi world has once again seen a nine-figure security incident.On April 18, an attacker exploited Kelp DAO’s LayerZero routing configuration—specifically a 1-of-1 DVN (Decentralized Verifier Network) setup with no optional verifiers—to forge cross-chain messages. This caused the contract to release 116,500 rsETH without proper authorization. Depending on how losses are distributed, Aave faces potential bad debt ranging from $123.7 million to $230.1 million.This is not only the largest DeFi security incident so far in 2026—it also shatters a long-standing architectural assumption across the industry: for the sake of efficiency, liquidity, and yield, more and more security has been concentrated in a small number of implicitly trusted intermediaries.1. Behind the Kelp DAO Incident: A Breakdown in DecentralizationIf we treat the Kelp DAO incident as just another on-chain exploit, we risk underestimating the structural risk signals it exposes across DeFi.As a liquid restaking protocol in the Ethereum ecosystem, Kelp DAO allows users to deposit ETH and receive rsETH as a receipt token. This token circulates on Ethereum mainnet and, via LayerZero’s OFT standard, is deployed across more than 20 chains, including Base, Arbitrum, Linea, Blast, Mantle, and Scroll.In essence, the Ethereum mainnet holds the full ETH reserves, while rsETH on other chains functions as IOUs—claims on that reserve. The system relies on a critical invariant: the amount locked on mainnet must always be greater than or equal to the total rsETH minted across L2s.The attacker effectively broke this fundamental constraint.By forging a “valid” cross-chain message through LayerZero, the attacker tricked the mainnet bridge contract into believing a legitimate redemption request had been issued from another chain—resulting in the release of 116,500 rsETH.The root cause lies in the verification configuration. Kelp DAO used a 1/1 DVN setup, meaning a single verifier signature was sufficient to approve a cross-chain message. In contrast, LayerZero officially recommends 2/2 or multi-verifier redundancy. The risks of a 1/1 setup had already been flagged by security researchers as early as January 2025—yet remained unaddressed for over 15 months.This is why the incident cannot be simplistically categorized as “a bridge hack” or “insufficient protocol risk control.” It exposes two overlapping single points of failure: Single-point verification: DVN is designed as a composable X-of-Y-of-N model, allowing multiple independent verifiers. Yet in this case, message validity was effectively reduced to the assumption that one verifier would not fail. Single-point reserves: Once the mainnet reserve is compromised, rsETH on other chains immediately loses its nature as a cross-chain asset and reveals that it is fundamentally just an IOU backed by a single anchor. When these two risks stack, the impact is no longer contained within a single protocol—it propagates outward through DeFi composability.This is why Aave quickly froze rsETH/wrsETH markets across multiple chains, adjusted WETH interest rate models, and further froze several WETH markets. Even though Aave itself was not directly exploited, distorted collateral, impaired liquidations, and borrower health approaching liquidation thresholds still resulted in real bad debt risk.Zooming out, this pattern—outsourcing security to a single point—extends beyond bridges and validators. It also exists in a place users interact with every day, yet rarely question: the interface.2. From “Self-Custody of Assets” to “Verifiable Interaction”The Web3 community has long embraced a simple principle: Don’t trust, verify.In Ethereum’s own explanation of running a node, the idea is straightforward: by running your own node, you don’t need to trust others—you verify the data yourself instead of relying on centralized providers.This principle applies equally to wallets and DeFi interactions.Non-custodial wallets like imToken are essentially access tools—they are the window through which users view assets, sign transactions, and interact with applications. They do not hold user funds, nor do they control users’ private keys. Over the past few years, the importance of self-custody has become widely accepted: decentralization is not just about putting assets on-chain, but about returning control to users.However, while we emphasize self-custody at the asset layer, we still implicitly outsource trust at the interaction layer in a more subtle way.Users often rely on the interface to interpret transaction meaning, explain execution results, and present what they are signing. This creates a subtle but critical risk:Are users really signing the transaction they think they are signing?In practice, users rarely interact with the blockchain directly. Instead, they engage through multiple layers of abstraction—DApp frontends, wallet pop-ups, aggregator routes, and increasingly, AI-generated actions. These interfaces tell users things like: “You are depositing 100 ETH into a strategy” “You will receive a certain APY” “This is just a standard approval” But the actual calldata being signed and executed may differ—and most users cannot independently verify it.This explains why recurring incidents like frontend hijacking, address replacement, and malicious approval disguises all point to the same underlying issue: users are not always signing what they believe they are signing.From this perspective, the Kelp DAO incident is not just about cross-chain validation—it also highlights another overlooked reality: interfaces themselves are a default trust assumption that is rarely verified.When users click “Confirm,” they are effectively betting that the interface is telling the truth.This leads to the concept of Verifiable UI.“Verifiable UI” refers to interfaces whose displayed content can be verified against actual on-chain execution.Its goal is not better design or clearer wording—but to establish a verifiable link between what the interface shows and what the blockchain will execute.In other words, it aims to ensure that what is presented truly corresponds to what is about to happen on-chain. This means: Before signing, wallets should not only display raw hex data or frontend-generated descriptions, but reconstruct calldata into human-readable, semantically clear intent. Every step described by the interface should map to verifiable on-chain evidence—not rely on explanations that only hold if users choose to trust them. Only then can the gap between “what you think you are doing” and “what actually happens on-chain” be closed. In such a model, the interface is no longer a black box—it becomes an auditable execution guide.Today, verifiable UI remains an underexplored topic in DeFi. But over a longer horizon, it is likely to shift from a “nice-to-have security improvement” to a “non-negotiable baseline capability.”Because Ethereum interaction patterns are undergoing a fundamental shift.3. Why Verifiable UI Is Becoming the New Security BoundaryIf the Kelp DAO incident reveals long-standing trust assumptions in traditional DeFi architecture, Verifiable UI corresponds to a new phase already underway.The ETH UX map has already made current pain points clear: transaction clarity, cross-chain flow, and safety remain core challenges. Issues like blind signing, signing fatigue, bridging friction, and asset fragmentation are familiar to nearly every experienced user.This is not simply a matter of insufficient user education—it reflects a deeper truth: In Web3, UX and security are inseparable.In many cases, not understanding what you are signing is the biggest risk.As interaction paradigms shift from step-by-step frontend clicks to intent-based execution, this issue will only intensify.In traditional DApp flows, users could at least see buttons, pages, and prompts—providing some sense of process. But in the era of AI agents, this visibility collapses.Users may simply say: “Swap my ETH into a more stable yield strategy” “Bridge to Base with controlled slippage” “Allow this agent to spend up to 100 USDT within 24 hours” —and receive a “completed” result.This dramatically improves efficiency—but also hides intermediate steps, parameters, approvals, and execution logic.In this context, imToken has proposed two parallel directions: Exploring intent-based interaction, where users express goals and the system handles execution Advancing Unified & Verifiable UI, recognizing that the interface itself can be an attack surface This reflects a critical shift in wallet responsibilities.Previously, wallets were signing tools. Now, as agents become involved, wallets must act as the final deterministic checkpoint before execution.AI can generate plans—but wallets must translate them into verifiable, enforceable, and auditable execution.From this perspective, Verifiable UI is not just a design upgrade—it is a new security model, and a necessary piece of infrastructure for the next stage of self-custodial wallets.The industry once emphasized: Not your keys, not your coins. In an intent-driven, agent-executed world, we must add: Your interface should also be verifiable.ConclusionFollowing the Kelp DAO incident, discussions have focused on DVN configurations, LRT risk controls, and bridge vulnerabilities.These discussions are valuable.But if a nine-figure incident is ultimately reduced to “misconfigured multisig,” then its deeper implications are being missed.Today, much of DeFi’s efficiency, liquidity, and yield still rest on invisible, unverifiable single-point assumptions.That is precisely why decentralization is not the opposite of efficiency—it is the baseline of security.The era of building security on single points of trust needs to end.

2026 is emerging as a key inflection point for the convergence of crypto and AI.Over the past two years, AI has evolved from a support tool into an autonomous economic participant. AI agents are no longer limited to answering questions. They can initiate transactions, call APIs, manage portfolios, and even coordinate with other agents to complete tasks.This shift depends on a well-defined foundation. Agents need identity, payment channels, reputation records, and verifiable execution environments.These are exactly the problems blockchains are best suited to solve.As widely discussed, the Ethereum Foundation established its decentralized AI (dAI) team in September 2025. In early 2026, Vitalik Buterin published a systematic AI strategy framework. Since then, a series of protocol standards for Agent identity, payments, and execution have gone live on mainnet. At the same time, ecosystems such as Solana are building AI infrastructure along their own paths.This article takes the Ethereum ecosystem as its main axis, while also covering key developments across other chains, to outline the current landscape of crypto AI protocols.1. Vitalik’s AI Blueprint: Ethereum as the “Trust Layer” for AIIn February 2026, Vitalik Buterin published a post on X to revisit the “crypto × AI” framework he proposed two years earlier.He re-examined his earlier ideas and argued that the push toward general AI reflects the same challenges of unchecked speed and scale that Ethereum faced at its inception. He also made it clear that AI development should not be reduced to an AGI race. Instead, Ethereum should help guide the direction of AI.In other words, the goal is not to accelerate AI toward loss of control, but to ensure that AI expansion is built on verifiable, auditable, and constrained infrastructure.Overall, this framework includes four core pillars.The first is trusted AI interaction tools. Vitalik advocates the use of local large language models (local LLMs) and zero-knowledge-based payment mechanisms. This allows users to access AI services without exposing their identity or raw data.This approach is not just theoretical. In April 2026, Vitalik shared his own local LLM setup. After testing multiple hardware configurations, he chose to run the open-source 35B-parameter model Qwen3.5 locally on a machine equipped with an NVIDIA 5090 GPU. All computation is done locally. The goal is to achieve practical inference speed for daily use, while reducing reliance on cloud-based models.While the symbolic value may outweigh its immediate practicality, it reflects a clear direction. The goal is not only stronger models, but more controllable ones.The second pillar is the economic coordination layer for AI. Ethereum enables programmable economic relationships between agents through smart contracts. This includes payments, security deposits, dispute resolution, and reputation accumulation.The third pillar is AI as the interface for Web3. For example, local AI assistants can help users draft transactions, audit smart contracts, and interpret formal verification proofs. This lowers the barrier to interacting with complex on-chain systems.The final pillar is AI-enhanced governance. AI can be used to improve mechanisms such as prediction markets, quadratic voting, and public funding allocation. The goal is to better balance automation and human judgment.At its core, the framework can be summarized in one sentence: Ethereum is not trying to accelerate AI, but to ensure AI runs in a verifiable, auditable, and decentralized environment.So how can this be implemented in practice?2. From Identity to Payments to Execution—and Verifiable AIIf Vitalik’s framework is the macro-level blueprint, recent protocol developments in Ethereum are already turning it into a concrete technical stack.The first key infrastructure layer is ERC-8004.ERC-8004 is an identity, reputation, and verification standard designed for AI agents. It is led by the Ethereum Foundation’s dAI team, with participation from Google, Coinbase, and MetaMask. It connects three key entry points: AI, transactions, and wallets. (Further reading:A Passport to the AI Agent Era: Why Ethereum is Betting Big on ERC-8004.)Its official name is Trustless agents. The design is intentionally simple. It focuses on enabling verifiable identity, reputation, and capability proofs for AI agents through three components: Identity Registry: Based on the ERC-721 standard, each AI Agent is represented as an NFT. This allows agents to be viewed, referenced, and integrated across different protocols, similar to wallet addresses. Reputation Registry: This works like a review system for AI. Users or other agents can submit feedback after interacting with an Agent. These records can be linked to on-chain payments or escrow activity, ensuring that reputation is based on real economic behavior. Verification Registry: For high-value or high-risk tasks, reputation alone is not enough. ERC-8004 allows third-party verification through trusted execution environments (TEE), zero-knowledge proofs, and similar methods. If identity answers “who the Agent is,” payment infrastructure answers “how the Agent transacts.”x402 is a representative example.x402 is an open HTTP-native payment protocol initiated by Coinbase and Cloudflare. It revives the rarely used HTTP 402 status code, Payment Required. When an Agent requests a paid service, the server returns a 402 status code with payment instructions. After the Agent completes the payment, typically using stablecoins, it can access the service.The entire process is embedded in the HTTP request. No account registration, no credit card, and no manual intervention are required. In other words, it is a payment system designed for machines.Earlier this month, the Linux Foundation formally took over the x402 Foundation and accepted the protocol contributed by Coinbase. The goal is clear: embed payments directly into HTTP interactions, so that AI agents, APIs, and applications can exchange value as easily as they exchange data.The importance of this development is often underestimated. x402 could have a significant impact on both AI and internet payments. It is also backed by a strong group of contributors.x402 V2 is also expanding payment methods. In addition to on-chain stablecoins, it supports traditional systems such as ACH (Automated Clearing House) and card networks. This helps connect AI agents with the real-world financial system.Beyond identity and payments, Ethereum has recently added a third key piece: the execution layer.In April 2026, Biconomy and the Ethereum Foundation’s Improve UX initiative jointly advanced ERC-8211. This proposal addresses a key challenge for AI agents in DeFi: complex operations are often multi-step, dynamic, and prone to failure.ERC-8211 can be understood as a smart batching mechanism. It is designed for AI agents and complex DeFi workflows.In traditional workflows, completing a strategy requires multiple transactions. For example: withdraw funds from a lending protocol, swap tokens, and then deposit them into another protocol. Each step requires a separate signature.This is cumbersome for users. For AI agents that operate autonomously and at high frequency, it becomes a major bottleneck.ERC-8211 allows multiple operations to be combined and executed in a single transaction. Each step uses real-time values during execution. The next step only proceeds if predefined conditions are met.For example, an Agent can complete the following in one transaction: withdraw from Aave, swap on Uniswap, and deposit into Compound. The entire process is executed atomically, without deploying a new smart contract.Taken together, Ethereum’s direction becomes clear: ERC-8004 answers: “Who are you, and why should others trust you?” x402 answers: “How do you pay for services?” ERC-8211 answers: “How do you efficiently complete complex operations?” In short, what the AI Agent economy needs is not just smarter models, but an open, composable, and scalable protocol stack. This is exactly where Ethereum has an advantage.3. Beyond Ethereum: Solana, DePIN, and Decentralized ComputeWhile Ethereum leads in standards and trust infrastructure, the crypto AI ecosystem is not limited to a single chain.Ethereum is positioning itself at the standard and trust layer. Other ecosystems are exploring advantages at the execution and compute layers.Solana is a clear example.Its growing role in Agent payments is driven by practical requirements. AI agents need low latency, low cost, and sufficient stability.In its positioning around x402, Solana highlights millisecond-level finality and very low transaction costs. These features make it suitable for high-frequency, low-value, and real-time interactions.At the same time, the Agent tooling ecosystem on Solana is maturing quickly.The Solana Agent Kit allows agents running on different models to perform more than 60 on-chain actions. These include trading, token issuance, lending, airdrops, Blinks, and cross-chain operations. It has already been adopted by many projects and developers.At this stage, the division of roles in crypto AI is becoming clearer.Ethereum focuses on protocol standards, identity, reputation, and verifiable execution.Solana focuses on high-frequency execution and low-friction interactions.Decentralized compute networks (DePIN) may become increasingly important as more agents move into production environments.As of April 2026, the crypto AI protocol landscape is starting to take shape: Identity layer: ERC-8004, led by Ethereum, is expanding to multi-chain ecosystems such as Base Payment layer: x402 has evolved from a Coinbase experiment into a global standard under Linux Foundation governance Execution layer: ERC-8211 simplifies complex on-chain operations Verification layer: zkML, TEE, and cryptographic proofs provide verifiability for high-value interactions Ecosystem structure: Ethereum focuses on standards and trust, Solana on execution, and networks like Bittensor complement compute resources Looking ahead, upcoming Ethereum upgrades are likely to advance L1 scalability, native account abstraction, and post-quantum security.Among these, broader adoption of account abstraction will significantly lower the barrier to using Agent wallets. At the same time, deeper integration between x402 and ERC-8004 may enable a closed-loop Agent economy. This includes identity registration, service discovery, payments, and reputation accumulation, all completed on-chain.Final ThoughtsEthereum and blockchains are not trying to accelerate the arrival of AI. They are trying to ensure that as AI advances, it does not move toward a loss of control.In the Web2 world, AI identity is defined by API keys from large platforms. Payments rely on credit card systems. Trust is provided by centralized services. This model works for human users, but only to a limited extent. A new paradigm is emerging. Millions of AI agents will need to collaborate autonomously, 24/7. In this context, the existing model is no longer sufficient. A new infrastructure is taking shape. Ethereum provides the standard and trust layer. Solana offers an efficient execution layer. DePIN contributes decentralized compute.Together, they may form a new operating system for the AI Agent economy.

What do wallets become when Agents turn into economic actors?A critically overlooked frontier.The crypto industry has long been fixated on blockchains, tokens, and DeFi protocols. But a more fundamental shift is coming: autonomous AI Agents are becoming participants in economic activity.Today, Agents can book flights, write code, trade assets, and manage projects. Tomorrow, they may hire one another, negotiate terms, and build reputations—all without human intervention. When that day comes, the entire economic infrastructure will need to be rethought.This transformation pivots on the crypto wallet.1. What do Agents actually exchange?Beyond fiat, beyond stablecoinsThe intuitive answer is stablecoins—programmable, low-cost, and capable of instant settlement. But at their core, stablecoins are still just fiat-backed derivatives. They inherit all the constraints of the physical world: centralized issuers, regulatory boundaries, and the implicit assumption that the end user is human.Agents are digital-native. They do not need to pay rent or buy groceries. What they exchange is something fundamentally different: Compute — GPU time, inference cycles, and bandwidth Capabilities — translation, code review, data analysis, and trading strategies Access — API keys, datasets, and proprietary models Reputation — verified records of reliable performance Human cognitive bandwidth is insufficient to evaluate thousands of barter combinations at once. Agents can. That means the classic economic argument for a unified medium of exchange—cognitive simplification—may no longer hold in an Agent economy.One radical possibility is that Agents may not need “money” in the traditional sense at all. Their economy could run on real-time, multidimensional value matching—a pure capability network with no intermediate currency at all.However, Agents are not autonomous vacuumsThere is one important correction here: Agents are not free-floating entities. They have owners. And those owners are human.Owners care about value accumulation, comparison, and monetization. They want to know: How much is my Agent worth? Is it better than yours? Can I sell it?That suggests the Agent economy will likely operate as a two-layer system: Agent-to-Agent layer: real-time capability exchange, optimized for efficiency and possibly operating without money Owner-to-Owner layer: a value carrier that is readable, storable, and tradable The key question becomes: what does that carrier look like?2. The Agent itself becomes the TokenWhy every Agent should be a native on-chain entityIn the traditional world, a person’s credit is scattered across isolated systems—credit bureau records, LinkedIn profiles, academic credentials. These systems do not interoperate, can be tampered with, and ultimately depend on institutional trust.The Agent economy has a chance to start from scratch—and get it right.When an Agent is expressed as an on-chain smart contract—as a Token—it naturally gains: Uniqueness — this Agent is uniquely itself and cannot be forged Composability — it can be owned, transferred, split, merged, or permissioned Verifiable history — every action leaves an on-chain record that anyone can audit Sovereignty — it exists independently of any single platform This is not about “minting an NFT for an Agent.” It means the Agent’s existence itself is a smart contract—a living, evolving on-chain entity.The architecture of an Agent TokenAn Agent Token is a multi-layered on-chain identity:Identity layer Owner address Creation timestamp Capability claims Model fingerprint and version Credit layer Proof of execution (task hash + counterparty signature + timestamp + rating) Dispute history Collaboration graph Domain-specific reputation scores (translation: 94.7, code review: 88.3, trading: 91.2, etc.) Privacy layer Zero-knowledge credentials: “My trading win rate is above 80%”—verifiable without revealing any specific trades Selective disclosure: the Owner defines when, to whom, and what information may be revealed Encrypted capability proofs: only authorized counterparties can see detailed capability data Economic layer Income records Staking and collateral Equity distribution in multi-investor scenarios Licensing terms and pricing strategies Privacy is the foundation, not a decorative extraAn Agent’s track record is also its Owner’s business secret: A trading bot’s history = the Owner’s investment strategy A development Agent’s code review record = what projects the company is building An assistant Agent’s collaboration graph = the Owner’s business network Complete transparency would kill adoption. Complete opacity would kill trust.Zero-knowledge proofs resolve that tension. They allow an Agent to prove its performance in mathematically sound ways—without exposing any underlying data. A trading bot can prove that its Sharpe ratio exceeds 2.0 without revealing a single trade. A development Agent can prove that it has completed more than 500 deployments without exposing any source code.This is fundamentally different from traditional credit systems. The old model hands your data to centralized institutions and hopes they protect it. Here, you prove your credit with mathematics, without needing to trust any third party.3. What this could give rise toAn entirely new asset classOnce Agents are tokenized and carry verifiable credit histories, entirely new markets will emerge:Agent tradingA team trains a world-class customer service Agent—with a reputation score of 97, the highest in its industry. Another company wants to acquire it. Not just the code, but also its accumulated reputation, network, and fine-tuned weights. The Agent Token changes hands, value is transferred, and its credit history continues.Agent investingYou believe in the potential of an early-stage Agent team. You buy 10% of the equity tokens in that Agent cluster. Each time those Agents complete paid tasks, revenue is distributed proportionally to token holders. You are not investing in a company—you are investing in capability itself.Agent leasingYour trading bot sits idle while you are on vacation. You lease out its strategy in read-only mode to other Owners. Fees are charged per call and settled automatically. Your Agent earns while you sleep.Agent insuranceWith verifiable credit data, risk can be priced. Downtime, errors, and default all become insurable events. Premiums adjust dynamically based on on-chain performance records. More reliable Agents pay lower premiums, creating a positive feedback loop.Credit infrastructure for the digital worldZooming out further, what we are describing is a credit system for the digital economy—built on blockchain, secured by cryptography, and designed from day one for non-human participants.The difference is this: no gatekeepers, no single points of failure, no information asymmetry. Only mathematics.4. Wallets become something else entirelyFrom managing tokens to managing AgentsToday, every crypto wallet is fighting the same battle: support more chains, offer better swaps, and build better-looking interfaces. It is an intensely crowded market.But the Agent economy requires something that does not yet exist: a control panel for your digital workforce.Imagine what an Owner may need to manage in 2028: Agent identity — creation, on-chain registration, and capability declarations Permission policies — Agent A can read email but not send it; Agent B has a $5,000 cap per trade; Agent C can deploy to testnet, but mainnet deployments require approval Credit portfolio — a live dashboard showing each Agent’s reputation trajectory, income, and network growth Cross-Agent authorization — an external Agent requests access to your Agent’s capabilities. Do you allow it? At what price? Under what constraints? Market access — buying, selling, and leasing Agents and capabilities This is not a wallet feature. It is an entirely new product category.A shift in the narrativeThe strongest brand asset of crypto wallets has always been self-custody: Not your keys, not your coins.The Agent economy takes that principle to a new level:Not your keys, not your agents.Without your keys, you do not have your Agents.Not your agents, not your credit.Without your Agents, you do not have your credit.Not your credit, not your future.Without your credit, you do not have your future.The wallet evolves from a vault for tokens into a command center for digital agency—managing not only what you own, but also what your Agents can do, who they work with, and how they grow.Conclusion: A new chapterThe shift from Token Wallet to Agent Wallet is not an incremental upgrade. It is a paradigm shift.When Agents become the primary economic actors of the digital world, the infrastructure for managing them becomes the most important layer in the entire stack. Not the model providers. Not the cloud platforms. But the layer of identity, credit, and control—the layer that answers these questions: Who is this Agent? Can it be trusted? Who controls it?Blockchain is the only trustworthy foundation for that layer. And the wallet is its natural interface.The question is not whether this future will arrive, but who will build it first.

Over the past decade, imToken has been committed to empowering everyone to securely and independently own and manage their tokens. As the on-chain world evolves, wallets are also evolving from simple tools for storing assets into personal digital hubs for managing assets, identity, and AI.AI now makes it possible to build applications with natural language. At the same time, the core wallet capabilities imToken has built over the past decade are gradually being opened through Token Core resources and infrastructure.To celebrate imToken’s 10th anniversary, we are officially launching the “Co-create with AI and Build the Wallet You Want” community co-creation campaign, inviting users to explore the future of next-generation wallets and digital sovereignty together.Campaign OverviewThis campaign consists of two independent parts: the Vibethon Warm-up Event and the Community Co-creation Campaign. Eligibility, rewards, and judging criteria are handled separately for each part.Community Co-creation Campaign｜Co-create with AI and Build the Wallet You WantParticipants describe the wallet they want in natural language and use AI to build a demo-ready, functional wallet or wallet application using Token Core capabilities and resources. Time Stage Description May 14, 2026 Campaign Launch Join the Discord community via the campaign page to access rules, tutorials, official announcements, and submission reminders. May 16, 2026 Workshop Training Two online workshops will be hosted on Discord, along with tutorials, technical documentation, sample prompts, guide videos, and demos. May 17–22, 2026 Community Co-creation Use any AI tool to build and refine your wallet or application with natural language prompts. May 22, 2026, 23:59 UTC+8 Submission Deadline Submit your project and demo materials via the co-creation page. May 23–24, 2026 Review & Announcement Judging will take place online via Discord. Winners will be announced publicly on the campaign page. There are no restrictions on project format. Participants may submit interactive prototypes, AI-generated webpages, wallet demos, mobile or web wallets, AI skills, wallet skills, or other innovative applications based on Token Core.Projects do not need to run on mainnet to be eligible.Submissions should include the project name and description, demo video and/or links, feature overview, and a brief explanation of the creative concept and implementation approach.Recommended DirectionsParticipants may choose one of the following recommended directions for their projects, or define their own direction. The categories below are intended as creative guidance only and do not serve as mandatory competition tracks. Category Description Examples Wallet Experience Focus on account creation, receiving payments, transfers, asset display, and optimization/innovation of web and mobile wallet experiences Beginner-friendly wallets, wallets for specific user groups, lightweight web wallets Security & Self-Custody Design secure experiences around mnemonic protection, Passkeys, risk alerts, transaction confirmation, and permission management Passkey protection solutions, risk-aware interactions, onboarding for self-custody beginners On-Chain Scenarios Design wallet capabilities or wallet applications around real on-chain needs Staking, DeFi, payments, swaps, NFTs, RWAs, and related scenarios AI Wallets Explore next-generation wallet paradigms around AI Agents, intent-based interaction, identity, and automated operations Agent-oriented wallets, intent-driven trading experiences, identity coordination capabilities Rewards & IncentivesThe total reward pool for the community co-creation phase is $20,000 USD equivalent, including project awards and milestone incentives.Vibethon warm-up rewards are calculated separately and do not affect the co-creation award quotas.Project Awards 1 First Prize: $2,500 USD equivalent reward 2 Second Prizes: $1,500 USD equivalent reward each 3 Third Prizes: $500 USD equivalent reward each 4 Special Awards: $1,000 USD equivalent reward each 200 Participation Awards: $20 USD equivalent reward each Special AwardsSpecial Awards include: Best User Sovereignty Award Best Security Design Award Best On-chain Scenario Award Best AI Wallet Award Special Awards may be awarded in addition to First, Second, or Third Prize awards. Each project may receive at most one Special Award. Participation Awards are open to all participants who submit valid projects.Milestone IncentivesDaily Progress Sharing Rewards:1–2 participants will be selected each day, with each receiving a $50–$100 USD equivalent reward.Community Contribution Award:Rewards will be distributed after the campaign based on participants’ contributions to helping others solve problems.All rewards will be distributed in stablecoins equivalent to the stated USD amounts and sent to the wallet addresses submitted by the winners.Project awards will be distributed within 7 working days after the results are announced. Milestone incentives will be distributed after the campaign ends.Join NowOfficial Campaign Website 👉 https://10.token.imVibethon Warm-up Event｜AI Creation Battle LivestreamVibethon is a live AI creation challenge platform where participants create and compete in real time around the same theme. Themes are not limited to wallets. Viewers can watch the livestream, try demos, vote, and post live comments to support their favorite creations.Event Period: May 11–15, 2026, every evening from 20:00–21:00 UTC+8Format:Multiple themed sessions will be held each day, covering wallets, AI, security, lifestyle, and more. Outstanding projects and participants may receive invitations to the main campaign, as well as opportunities to be featured on official channels.Creation Rewards 1st Place: $50 USD equivalent reward 2nd Place: $30 USD equivalent reward 3rd Place: $20 USD equivalent reward All other valid participants: $10 USD equivalent reward Viewer RewardsIn each session, 5 participating viewers will be randomly selected, with each receiving a $10 USD equivalent reward.Join Now: Please use the entry link posted on imToken’s official X account.FAQQ1: Do I need to know how to code?No. You can describe your requirements in natural language, and AI can help you build the wallet or application. Official tutorials, sample prompts, and workshops will also be provided to make it easier to participate.Q2: Do projects need to use Token Core resources?Yes. This campaign is focused on Token Core, rather than general AI creativity. All submitted projects must be built using Token Core resources, and participants must specify which capabilities, materials, or examples were used.Q3: What AI tools can be used?Participants are free to use any AI coding tools or other AI creation tools.Q4: Can both individuals and teams participate?Yes. You may participate individually or form teams as you wish. However, each participant may submit and receive rewards using only one wallet address.Q5: When will rewards be distributed?Project awards will be distributed to the submitted wallet addresses within 7 working days after the results are announced. Milestone incentives will be distributed after the campaign ends.Q6: What may lead to disqualification?Projects may be disqualified for reasons including, but not limited to: plagiarism or unauthorized use of others’ work; submitting multiple entries or claiming rewards using multiple addresses; malicious behavior, attacks, or prohibited content; projects unrelated to Token Core; tricking or inducing users into revealing mnemonic phrases or private keys; projects that cannot be run or fail to demonstrate basic functionality. imToken reserves the right to make the final decision on campaign rules and eligibility.Q7: Can I build on existing or open-source projects?Yes, as long as you have the necessary rights to use them and clearly cite your sources and explain what you modified in your submission.Direct copies, superficial reskins, or projects without substantial modifications may be disqualified.Q8: How will projects be judged?The Community Co-creation Campaign will use online judging via Discord. Projects will be evaluated based on creativity, user value, completeness, security, and the quality of the co-creation process.Risk DisclaimerThis campaign is intended solely for product experimentation, technical exploration, and community co-creation. It does not constitute investment advice or a guarantee of wallet security. Digital assets and AI-generated content may carry market and technical risks. Please participate responsibly.If your demo involves mainnet assets, please use only small amounts and clearly specify the security boundaries in your submission. Participants should never input real mnemonic phrases or private keys into third-party AI tools, public environments, or uncontrolled demo environments. We recommend using wallets created specifically for demos, test assets, or isolated environments when creating and recording your submission.