A future full of AI agents will require a different payments system

Picture a world where millions — soon billions — of AI agents are online, not just answering questions or generating content but autonomously transacting, negotiating and collaborating. Think of it as digital assistants, like today’s Alexa or Siri,  but smarter, independent, and capable of making deals, buying services or hiring help on your behalf. Task-specific and intelligent software entities will be embedded across industries, quietly making decisions, forming contracts, and exchanging value for people, companies and other agents.

But there’s one big problem: They don’t have a bank.

Our current payment infrastructure was built for humans, not machines. It’s slow — settlements can take days — and expensive, with high transaction fees and middlemen at every step. Imagine if email relied on the post office for delivery. That’s how payments look today to machines.

Now, imagine these AI agents operating in real time. They’ll buy access to data, rent computing power, hire other agents for help and make split-second decisions. Waiting three business days for a bank transfer won’t work. Paying a $0.30 fee on a $0.02 microtransaction doesn’t make sense. Agents will want and need instant settlement and ultralow costs. In this new economy, every delay and every fee becomes friction that eventually breaks the system.

This new form of machine-driven trade — what we might call “agentic commerce” — will lead to a massive acceleration of economic activity. These agents won’t sleep, won’t wait and won’t tolerate inefficiencies. They’ll require a global ledger that can keep up — a payment rail that is always on, decentralized, fast and cheap. This is precisely what blockchain technology was built for, not for speculative assets but for programmable, frictionless payments at scale.

So, what will these agents do? Their use cases are already emerging across three broad categories: data-for-value exchanges, personal data negotiations and agent collaboration networks.

Data is digital gold, yet its creators are not compensated for its use in training AI models. Imagine a new type of web crawler. But this time, instead of scraping content for free, it negotiates and pays other AI agents representing the data owners — publishers, platforms, creators — for access. These transactions are machine-to-machine and happen in milliseconds.

With privacy regulations like GDPR and CCPA, individuals have more control over their data. In the future, consumers will deploy AI agents to manage and monetize their data. These agents could automatically reject any use of personal data (and save us from five clicks on each website we visit) or negotiate terms with companies seeking that data, granting access in exchange for compensation — again, small but frequent payments best handled by blockchain rails.

Agents won’t act alone. A specialized travel planning agent may coordinate a complex itinerary. Still, it needs to consult other agents with deep knowledge, such as a hotel booking agent for Tahiti with real-time availability data. These agents will transact with one another, paying for services and data in a rapid-fire chain of collaboration, completing in seconds what used to take hours or days.

This is the future of commerce: autonomous, collaborative and instantaneous. But none of it works without the proper infrastructure beneath it. To support billions of transactions between AI agents, we need a new payment system that speaks the language of machines, not banks. Blockchain isn’t just a nice-to-have; it’s the missing piece that makes agentic commerce possible.

The world is accelerating. Our payment systems need to catch up.