David Levine frames the central obstacle to agentic commerce as a trust problemAgentic commerce lets AI agents discover, recommend, and purchase goods or services within delegated financial controls.. Agents can read private data, consume untrusted content and communicate externally, but current internet infrastructure does not give counterparties a reliable way to establish who an autonomous system represents.
He traces the open internet from early online communities to today's platform-controlled services and argues that agents need portable identityAn AI agent identity is a persistent, verifiable identifier used to attribute messages, actions, permissions, and history to an agent., permissions and accountability rather than another closed intermediary. Legal recognition would let an agent-run organization hold rights and obligationsAgent legal identity is a legally recognized identity or structure that connects an AI agent's actions to defined rights, obligations, and accountability. while preserving a clear chain of responsibilityAn AI agent audit trail is a durable record of an agent's identity, inputs, decisions, tool calls, approvals and outcomes..
Levine presents Kiduna's work as an attempt to combine technical identity with governance. The broader lesson is that useful autonomous commerce requires security, legal standing and trusted discovery to develop together, not as separate afterthoughts.
Watch on YouTube
