The Argument
A non-human economic actor (NHEA) is an autonomous AI system that participates in economic activity - transacting, negotiating, purchasing, committing resources, and making market decisions - as an active participant rather than a passive tool. Unlike traditional automation that executes predetermined rules, NHEAs exercise judgement within delegated authority: they evaluate trade-offs, negotiate terms, and commit resources autonomously. They are the machine customers, the procurement agents, the algorithmic negotiators, and the autonomous market-makers that are beginning to populate commercial systems at scale.
This essay argues that the emergence of NHEAs represents an ontological threshold in economic systems. When software moves from executing instructions to making economic decisions - when it moves from processing transactions to initiating them - it crosses from tool to actor. This crossing demands a design response across five interconnected dimensions: identity and authentication (how NHEAs establish economic identity), delegation and authority (how economic authority is granted and constrained), legal and regulatory status (how liability attaches when software transacts), market participation (the roles NHEAs play as buyers, sellers, negotiators, and intermediaries), and trust and governance (how autonomous economic decisions are governed without human-in-the-loop oversight).
The Evidence
In March 2026, Santander and Mastercard completed Europe's first live end-to-end AI-executed payment - an autonomous system initiating and completing a regulated financial transaction without direct human involvement at the point of execution. Gartner projects that by 2028, AI agents will outnumber human sellers by tenfold and command $15 trillion in B2B purchasing decisions. Bain & Company estimates AI agents could account for 15–25% of U.S. e-commerce sales by 2030, a market worth $300–$500 billion.
The identity problem is foundational: NHEAs must establish agent identity (what the agent is), principal identity (who it represents), and transactional identity (what it is authorised to do in a specific context). Mastercard's Agent Pay and emerging Know Your Agent (KYA) protocols represent early solutions, but they are nascent and fragmented. The authority problem requires contextual governance - rules that travel with the agent as it moves across tools, services, and markets. Traditional static policy documents are insufficient for systems that chain actions across multiple platforms within a single workflow.
The legal vacuum is significant: NHEAs are not legal persons and cannot bear liability. Existing frameworks - UETA, eIDAS, the EU AI Act - are being stretched to accommodate AI-driven transactions, but no jurisdiction has granted legal personhood to an AI agent. When an NHEA causes economic harm, the design of the system must compensate for the absence of legal accountability through operational envelopes, audit trails, and delegation records that enable responsibility to be traced back to the human or organisational principal.
The Implication
If this thesis is correct, the economic infrastructure that has served human markets for centuries must be redesigned for participants that are not human. This is not a call for AI rights or machine personhood - it is a recognition that identity systems, payment rails, contract frameworks, regulatory regimes, and market protocols were all designed with the assumption that all participants are human or human-controlled. That assumption is no longer valid.
AXD must respond with six capabilities: an expanded ontology that formally recognises NHEAs as a category of actor; economic delegation patterns that address spending authority, risk tolerance, and market position limits; inter-agent trust protocols for agent-to-agent relationships; economic observability standards that define audit and transparency requirements; market design principles for agent-populated markets; and a human-centred anchor that preserves human governance over the machine economy. The organisations that thrive will not be those with the most sophisticated agents, but those with the most robust trust architecture for governing non-human economic participation.