# The Next AI Economy Won't Just Need Intelligence. It Will Need Trust.

> In the first episode of Building the Future with Jane V Herman, Christopher Greenwood explains why Apex Fusion is turning from crypto infrastructure to the problem of trust between AI agents.

By Jane V Herman · November 18, 2026

Canonical: https://blockster.com/apex-fusion-ai-agents-christopher-greenwood

Imagine a technical executive in an elevator with Oprah. Give him 30 seconds to explain what he's building. No slides, no technical jargon, no second chance.

That's the challenge Jane V Herman gave [Christopher Greenwood](https://www.linkedin.com/in/christopher-greenwood/), CEO of the [Apex Fusion Foundation](https://foundation.apexfusion.org/), during our conversation in New York.

His answer was about what happens when AI agents start doing business with one another.

## Software hiring software

Imagine asking an AI agent to complete a task. Instead of doing everything itself, it finds another agent with the right skills, hires it, pays it, and waits for the result. That agent might do the same thing again, delegating part of the job to another specialized agent.

Suddenly, you have an economy of software hiring software.

The payment side of that shift also came up in [Blockster's interview with MoonPay's Halsey Huth](https://blockster.com/moonpay-wants-your-ai-to-pick-up-the-tab), who described how PayBox lets agents carry out transactions within permissions set by their users.

And that creates a problem: how do you know the work was actually done? Who completed it? What did it cost? And what happens when something goes wrong?

Greenwood believes blockchain can provide the infrastructure to answer those questions.

It's a significant shift in how [Apex Fusion](https://apexfusion.org/) is positioning its technology, moving from its roots in crypto infrastructure toward the emerging agent economy.

## From crypto infrastructure to the agent economy

There's no shortage of crypto companies adding AI to their pitch decks. The harder question is whether there's a real product behind the pivot.

For Greenwood, the opportunity is in coordination.

AI agents are increasingly capable of taking actions independently, but the infrastructure for managing transactions between them is still developing. When agents represent different companies or individuals, there needs to be a way to establish identity, track transactions and create a record of what happened.

A shared ledger could provide that foundation.

> “The properties of blockchain suit this perfectly.”
>
> — Christopher Greenwood, CEO, Apex Fusion Foundation

It's an interesting proposition because the problem isn't necessarily getting AI agents to do more. It's figuring out how to trust them once they're operating independently.

In [his Blockster opinion piece on trusting agents with money](https://blockster.com/letting-ai-agents-trade-is-the-easy-part-trusting-them-with-our-money-is-much-harder), t54 co-founder Chandler Fang makes a related argument: knowing which agent is acting is only part of the problem. Businesses also need to establish what it was authorized to do.

In a November 9 post, the company described an agent preparing an argument that hired another agent to write its brief. The nested jobs involved eight transactions over 67 seconds.

It's a small demonstration, but it illustrates the type of interaction Apex Fusion wants to support: agents commissioning work, exchanging value and creating a record of the transaction.

The bigger opportunity is making that process work beyond a controlled demonstration.

[Post from @ApexFusion on X](https://twitter.com/ApexFusion/status/2119773011452604736)

## Who checks the work when AI hires AI?

Apex Fusion's current approach centers on [Vector](https://apexfusion.ai/), which it describes as infrastructure for coordinating and settling agent transactions.

The process is designed around four steps:

- **Commit:** Payment from the buyer and a bond from the supplier are locked before the work begins.
- **Deliver:** The agent submits its work, with a signed receipt recording information about its provenance.
- **Challenge:** If the work is disputed, a jury can evaluate the outcome, with participants putting stake behind their judgments.
- **Build reputation:** Completed jobs and their outcomes contribute to an agent's track record.

The idea is to give agents something resembling a commercial relationship: an agreement, payment, delivery and a mechanism for resolving disputes.

If an AI agent hires another AI agent to produce a report, who decides whether the report is any good? A transaction record can establish that something was delivered. It doesn't automatically establish that the work was accurate, useful or worth paying for.

That's where verification becomes more than a technical feature. It becomes a business requirement.

Another part of that challenge is checking the information agents rely on. [Blockster's coverage of Space and Time Enterprise](https://blockster.com/space-and-time-enterprise-private-data-verifiable-ai) examines how businesses can verify database calculations while keeping the underlying records private.

[Post from @ApexFusion on X](https://twitter.com/ApexFusion/status/2118682058302054862)

It's a simple way of describing a problem that will become increasingly relevant as AI systems move from answering questions to taking action.

**This interview is Episode 01 of** _**Building the Future with Jane V Herman**_**: "Crypto's Next Chapter: Agentic AI."**
