Original: Deep Tide TechFlow

On January 28, the Ethereum Foundation officially announced that the ERC-8004 protocol is about to launch on the mainnet.

We already mentioned this in an article we published last October. If you don’t know much about it at all, you can refer here: 《x402 The market is gradually getting more and more cutthroat—seizing new asset opportunities inside ERC-8004》

It actually has an official name: “Trustless Agents”—trustless agents. Translated into plain human language, roughly it means:

Give AI agents an on-chain ID.

The Ethereum Foundation rarely puts this much effort behind pushing an ERC standard. They specifically set up a team called dAI, added ERC-8004 to the 2026 strategic roadmap, and drafted it together with Google, Coinbase, and MetaMask. In November, they also held a Trustless Agents Day at DevConnect to drum up momentum.

The last time Ethereum pushed standards this seriously was ERC-20 and ERC-721.

One defines a token. One defines an NFT.

So now it’s AI’s turn?

Ethereum’s AI anxiety

Why is it so urgent?

Take a look at some data. According to Cookie.fun’s statistics on the market cap distribution of AI Agent tokens, Solana and Base combined account for 96%. On Ethereum mainnet, there are only a handful of AI Agent projects that anyone can name—well within the fingers of one hand.

On Ethereum mainnet, there are only a handful of AI Agent projects that anyone can name—well within the fingers of one hand.

In April 2025, the ETH-to-BTC exchange rate fell to 0.017, a five-year low. Back then, people said Ethereum wasn’t the future.

During the DeFi boom, Ethereum was the home field. During the NFT boom, Ethereum was also the home field. When the AI Agent boom kicked off, the home field changed hands.

Solana processes 36 million transactions per day, while Ethereum mainnet processes 1.13 million. High gas fees and slow speed mean developers are “voting with their feet.” When Virtuals Protocol launched on Base, earlier ai16z chose Solana as well, and even Coinbase’s own AI project wasn’t placed on Ethereum mainnet.

Ethereum needs a new story.

ERC-8004 may be the beginning of that story.

Let’s recap ERC-8004

Back to the standard itself.

How exactly does ERC-8004 do this whole “issuing an on-chain ID to AI agents” thing?

You don’t need to understand any technical details—just know there are three registries.

The first is the identity registry. Built on ERC-721, each AI agent mints an NFT to prove: “I am who I am.”

The second is the reputation registry. It records the agent’s historical performance—who has used it, what they thought of it, and whether it has done anything bad.

The third is the verification registry. It lets third-party organizations “stamp” and endorse the agent—such as: “This agent passed a certain security audit.”

Put the three together, and they solve one problem: when two AI agents meet on-chain, how do you know whether the other one is trustworthy?

In the past, the answer was: you don’t know—you can only rely on people. ERC-8004’s answer is: check the on-chain records.

This set of ideas wasn’t created by Ethereum itself.

Its underlying logic comes from Google’s A2A protocol published last year—Agent-to-Agent—so that AI can communicate and call each other. ERC-8004 builds on that by adding another layer:

Trust backed by the blockchain.

Google’s A2A solves the communication problem; Ethereum’s ERC-8004 solves the trust problem. One manages to talk, and one verifies who you really are.

Is issuing IDs a good business?

Boldly speculate: Ethereum’s logic might be something like this:

For AI agents to be truly useful, they need to be able to manage money themselves. Not posting on social media, not chat—directly operating on-chain assets. Signing transactions, calling contracts, cross-protocol arbitrage…

Right now, nobody dares to do this at scale. The reason is simple: how do you know the agent won’t just move your money away? The ClawdBot that’s been exploding in popularity these days already has community users posting related negative incidents.

Web2’s solution is platform endorsement. If you use OpenAI’s API, you trust OpenAI. If something goes wrong, you go after OpenAI.

Web3 doesn’t have that. Agents are open source, deployment is permissionless, and there’s no one watching over them on-chain. If you call an unknown agent’s service, you can’t easily find out who’s behind it, whether the code has issues, or whether it has a history of doing harm.

To put it bluntly, ERC-8004 essentially brings traditional finance’s KYC process onto the blockchain. And Ethereum is betting that once AI agents start touching real money, this will become a must-have.

For DeFi protocols to integrate external agents, they first need to check the agent’s on-chain identity. For institutions to use agents for transaction execution, they first need to review its historical records. Audit firms can issue on-chain certifications to agents—like performing security audits for smart contracts.

This is a competitive “grabbing the position” move.

Ethereum knows it already lost on the execution layer, but nobody has occupied the trust layer yet. Institutional recognition, the security audit ecosystem, and TVL scale—these are Ethereum’s existing assets. ERC-8004 packages these assets into a standard, and rushes to define what “compliance for AI agents” looks like before anyone else does.

But the question is: does this need exist right now?

Standards come before demand

After laying out Ethereum’s算盘, let’s face reality: what are on-chain AI agents doing now?

After last year’s AI meme wave burned through, and with several leading AI companies making dramatic progress with AI products over the past one or two years, not many people are still paying attention to on-chain AI agents.

But they have still made progress.

For example, ai16z has already changed its name to ElizaOS—evolving from a single agent into a cross-chain platform. Virtuals Protocol is building an AI DAPP and plans to move into physical robotics in 2026. Other AI agents, such as those in Surf, can also automatically execute DeFi trading strategies.

But here’s the problem: do they really need ERC-8004?

Luna’s users trust Luna because it’s made by Virtuals’ core team. Agents on ElizaOS are used because they run within the ElizaOS framework. If Surf helps you execute strategies, many times it’s because you trust the application itself.

Trust comes from the platform, not from an on-chain identity.

The scenario ERC-8004 envisions is: a stranger agent comes to you. There’s no platform endorsement, no brand awareness—so you can only judge whether it’s trustworthy by looking at on-chain records.

When would this scenario actually happen?

When AI agents truly achieve autonomous calls across protocol boundaries, platform boundaries, and organizational boundaries. An agent borrows money from Aave, trades on Uniswap, then makes yield in another protocol—through the whole process, without human approval the entire time…

But this scenario doesn’t exist yet.

Today’s AI agents, no matter how complex their functions are, fundamentally still operate within a single platform. They don’t need to prove themselves to unknown protocols, because they literally won’t go knocking on unknown protocols’ doors.

Given the current heat in the crypto market, they also have no reason to go knock on each other’s doors—unless they can somehow work together to create a new narrative.

So ERC-8004 solves a future problem.

If AI agents move from being toys to being tools, Ethereum’s trust infrastructure will become valuable. If the scale of agent economics becomes large enough and cross-platform calls become the norm, ERC-8004 can collect tolls.

There are many “ifs.”

So in this future-oriented setup, the first ones likely to act are institutions.

At the end of 2025, SharpLink Gaming announced it would invest $170 million into an Ethereum restaking strategy. Around the same time, exchanges’ net ETH outflows exceeded 23k coins, flowing to private wallets and staking protocols.

This money may be buying Ethereum 12 to 18 months down the road.

For retail investors, in reality, ERC-8004 isn’t much of a good catalyst.

Bet on ERC-8004 itself? It’s an open standard with no token—you can’t invest directly; you can only look for some related small projects. Betting on Ethereum isn’t necessarily wrong either, but Ethereum’s price is influenced by too many factors, and AI agents are just one narrative among them.

So, right now there isn’t a clean underlying asset that lets you precisely bet on the thesis “AI agents need on-chain identity.”

Ethereum isn’t only AI infrastructure, and Ethereum’s identity anxiety won’t disappear just because AI fully resolves everything. The business of making AI IDs still has a long road ahead.

Welcome to join ChainCatcher’s official community

Telegram subscription: https://t.me/chaincatcher;

Official Twitter account: https://x.com/ChainCatcher_