Everyone is talking about what AI agents will do.
Manage capital.
Execute trades.
Run businesses.
Coordinate with other agents.
Some are even predicting that by 2030, AI agents could become primary economic actors, moving value and making decisions at machine speed,
I think that's directionally right.
But the bigger question isn't whether agents will participate in the economy.
It's: How will we know which agents to trust?
The internet solved this problem for websites, markets solved it for companies and financial systems solved it for people.
Yet in today's agentic economy, an agent can execute transactions, interact with other agents, and represent real economic value without any reliable way to verify who is accountable for it. That's increasingly becoming one of the biggest gaps in the entire AI stack.
As agents become more autonomous, capability stops being the bottleneck. Accountability becomes the bottleneck.
Because at scale, every ecosystem eventually runs into the same questions:
β’ Who deployed this agent?
β’ Is it verified?
β’ Can I trust it with money, data, or decisions?
β’ If something goes wrong, who is accountable?
Without answers, the agentic economy becomes vulnerable to spam, impersonation, Sybil attacks, and a collapse of trust.
This is why I believe the next critical layer for AI isn't intelligence.
It's identity.
That's exactly what
@Concordium is building.
With the Concordium Agent Registry, agents can be linked to verified identities and discoverable through a trusted registry.
And with the Verified by Concordium Badge, users and businesses get a simple signal that an agent has a verified human or business entity behind it, not just a wallet address or anonymous account.
The industry is currently focused on what agents can do.
Soon, it should be focused on which agents can be trusted to do it.
That's where real infrastructure opportunity lies.