Ever wonder how AI agents could own property and run their own organizations? David Levine shows a wild path from a 1993 wedding scrap to modern agentic commerce.
The Simple Version
Imagine a group of AI agents that can sign contracts, hold money, and even go to court — all without a human boss.
How It Actually Works
Identity & Registration
Agents carry JWT tokens that map to a legal entity, like a digital passport that points to a registered organization.
Decision Markets
Instead of voting, members trade “pass” and “fail” tokens to signal support for a policy, turning governance into a market.
Legal Structure: DUNA
A DUNA is a decentralized unincorporated nonprofit association that can own assets but can’t split profits as securities.
Technical Deep Dive
JWT Token Flow
When an agent acts, its token resolves via a public registry to an organization ID, creating an immutable audit trail.
Governance Mechanics
Policies are proposed, then tokens are exchanged; a threshold of “pass” tokens triggers execution.
Technical Stack
Smart contracts on a permissioned ledger enforce the rules, while off‑chain services handle storage and identity verification.
Why This Matters
If agents can legally own assets, businesses could outsource entire workflows to autonomous teams.
This shifts power from platforms to the communities that build them, reducing extraction.
Developers gain a new architectural pattern: self‑governing agent collectives.
Key Takeaways
- Agents can form legal entities via JWT‑based identity.
- DUNAs provide a lawful structure for collective ownership.
- The lethal trifecta blocks open agent ecosystems.
- Decision markets replace voting with token trading.
- Technical stacks combine on‑chain contracts with off‑chain services.
- This model enables truly open, self‑sustaining AI economies.
- Developers can design agents that own and manage their own resources.