# The battle for AI agents: Silicon Valley versus the crypto industry

> The battle for AI agents: Will Silicon Valley or the crypto sector dominate? We analyze the future of the new "agentic economy"
> By Covadonga Fernandez · AI Agents · 2026-03-20

Although the crypto industry has a short memory, its pattern repeats itself. First came the ICOs. Then DeFi and NFTs. Later, DAOs. Each cycle had a narrative capable of promising a radical transformation of the internet, finance or digital organization. Some left behind useful tools and concrete use cases. But almost none fulfilled, at least in the short term, the magnitude of the expectations they generated.

Now, the new great promise is called the "agentic economy," an economy in which autonomous artificial intelligence agents buy, sell, contract services and pay each other without human intervention. The idea has force and a lot of narrative. And, precisely for that reason, it is worth looking at it with some distance.

## Coinbase, Circle and Binance

Much of the current enthusiasm comes from players with very clear interests in making that scenario a reality. Brian Armstrong, from Coinbase, Changpeng Zhao, at Binance, and Jeremy Allaire, from Circle, argue that if AI agents are going to operate economically, they will need an internet-native payment system, and that system, according to them, will be crypto. Allaire has even stated that stablecoins could become the native currency of machine-to-machine commerce, a statement that places USDC, his stablecoin, at the center of that narrative.

The argument is not hard to understand. An AI agent does not fit well into the traditional financial architecture. It cannot open a bank account like a person, it cannot pass identification processes designed for humans and it does not have easy access to conventional banking infrastructure. It can, however, operate with a wallet, receive funds, execute payments and move across on-chain networks with much less friction. It is on that premise that companies like Coinbase, Circle and others are trying to build the financial layer of the future agent economy.

## Parallels with other crypto cycles

But one thing is that the logic is acceptable and another is that the end market will actually develop as they say. That is where the parallel with other crypto cycles appears. The sector's history is full of ideas ahead of their time or inflated by the commercial incentives of those who stood to benefit most from them.

[The bad company your AI agent keeps, the risk no one sees coming](https://observatorioblockchain.com/en/ai-agents/las-malas-companias-de-tu-agente-de-ia-el-riesgo-que-nadie-ve/)

Does that force us to ask whether the same thing that already happened with ICOs, NFTs or DAOs could happen with the agentic economy? That is, there may be a real technological base, but an excessive expectation about its speed of adoption, its immediate scale or its economic impact.

At present it is not clear to what extent there will be massive demand for transactions between agents beyond very specific niches. Nor is it resolved how legal liability, security, fraud, traceability or errors will be managed when the party operating is not a person but an autonomous system. And although the crypto thesis is highly seductive, it also cannot be taken for granted that the future of agents will be built exclusively on decentralized networks.

In fact, recent moves by the big technology companies point to something else. Meta announced the acquisition of Moltbook, a social network designed for AI agents, a sign that interest in this ecosystem no longer belongs only to the crypto world, but also to the big technology platforms.

## Will it need cryptocurrencies?

Something similar is happening at OpenAI. In February it hired Peter Steinberger, creator of OpenClaw, one of the most discussed open source projects in the recent narrative of personal agents. Sam Altman, CEO of OpenAI, explained that Steinberger would devote himself to driving the next generation of personal agents, while OpenClaw would continue to exist as an open project with OpenAI's support. Undoubtedly, a clear sign that the big labs are also absorbing talent and ideas in this direction.

All of this suggests that the underlying trend is real. Agents are ceasing to be a simple demo and becoming a strategic category. The question is not so much whether there will be an agent economy, but what form it will take, who will control the infrastructure and how much of what is promised today will survive the passage of time.

Another of the big unknowns is whether this agent economy will really need cryptocurrencies to exist. In practice, it could also develop on traditional infrastructures, with agents operating within platforms such as those of OpenAI, Meta or Google, using accounts linked to users and conventional payment systems. However, that model would be, by definition, closed and dependent on intermediaries.

## The battle for AI agents

The crypto sector's thesis points to agents with their own wallets, capable of operating autonomously and without permission on open networks. Along these lines, Coinbase is also advancing with digital identity, exploring solutions with Word to give agents verifiable credentials that allow them to interact securely in open environments.

The difference is not minor, because at bottom it defines whether it will be the big technology platforms that manage identities, payments and interaction rules, as happens today in closed ecosystems, or whether that control will shift toward open infrastructures where agents operate with greater autonomy.

But even in this second scenario, power does not disappear. Protocols also have developers, foundations and players with economic incentives that influence their evolution. In practice, what is at stake is the power struggle between the consolidated big technology companies, which seek to integrate agents within their own environments, and a new generation of blockchain-linked companies and projects trying to build an open economic layer from scratch.

Companies like [Coinbase](https://x.com/worldnetwork/status/2033923693529993334), Binance or Circle are not neutral observers. All three have obvious incentives to present wallets, stablecoins and on-chain payments as the natural solution to the problem. That does not invalidate their thesis, but it does force a distinction between analysis and promotion.
