Ethereum Co-Founder Joseph Lubin Predicts the Merger of AI and Crypto. Here's the Most Likely Scenario for Ethereum.

In this news:

Key takeaways

Joseph Lubin predicts a merger between artificial intelligence and cryptocurrency to establish verifiable trust.
While agent-driven economic activity remains largely theoretical today, adoption is growing across Ethereum and Layer-2 networks.

On Oct. 7, Ethereum co-founder Joseph Lubin said on CNBC that while AI might offer creativity and expertise, it can't create trust, which blockchains like his can help to supply. In his view, AI and crypto will thus end up merging, becoming the system where programmatic work can be commissioned, completed, and compensated.

There's a little bit of evidence that Lubin's world is already starting to peek into existence. Here's what it'd mean for Ethereum if that continues.

Ethereum is currently home to 9.2% of all registered AI agents

Today, one of Ethereum's biggest contributions to the AI-in-crypto segment is ERC-8004, a new standard for blockchains that calls for tracking the identities and reputations of AI agents. As of Oct. 20, Ethereum hosted 72,028 registered AI agents, about 9.2% of the 783,881 agents registered under ERC-8004.

It also holds the most capital, which agents will likely need to use, with $49.6 billion in capital stored in its decentralized finance apps, and more than $144 billion in stablecoins. Liquidity in DeFi services and stablecoins are both big parts of what makes a cryptocurrency valuable to agents moving money, as the size of those footprints determines how much capital is available for the ecosystem to use for services and products.

Are AI agents actually creating real value on Ethereum yet?

Today, the agent economy is largely theoretical.

Monthly agent payments over x402, a standard for paying in stablecoins, peaked at $20.5 million in November 2025, and then fell to $263,863 in September 2026. Meanwhile, Ethereum collected $17.7 million in transaction fees in the 30 days ending on Oct. 20. Agents supplied almost none of that, since Ethereum's main chain had settled just 644 agent payments as of Oct. 20.

Still, there are signs Lubin's vision is taking shape. Ethereum hosted about 32,343 registered agents as of May 13 of this year, so its count has more than doubled in five months.

Moving forward, expect its capital to attract more agents that are native to other networks, and expect more developers to launch their agents on Ethereum. Over the next five years, the sum of agents could explode, and with it, the frequency and amount of their transactions, as well as the diversity of their needs and offered services.

If that happens, holders might benefit, as agents on Ethereum and on Layer-2 networks like Base, which charges fees in Ethereum's coin, would pay fees in it. That makes Ethereum a practical way to add AI exposure to a well-balanced crypto portfolio, even if agent-driven fees take years to ramp up, which they might.

The most likely scenario is that Ethereum's capital pulls agents in gradually, with payments flowing through its Layer-2 networks.

Watch two numbers: Ethereum's share of registered agents, and monthly x402 payments. If agents are making payments on Ethereum and doing most of their business there, it'll be bullish for the coin even if the trust element that Lubin is hoping for never becomes as important as he expects.

Share the article: