Agent ID Standard v0.1

This document defines the first public version of the Agent ID standard for artificial agents.

1. Purpose

The purpose of this standard is to introduce a minimal, persistent identity layer for artificial agents.

Artificial agents increasingly act on behalf of humans and organizations, yet lack stable identifiers across systems.

Agent ID addresses this gap by defining a neutral reference mechanism for identifying agents over time.

2. Artificial Agent

An artificial agent is a software-based entity that:

• acts on behalf of a human or organization,
• performs tasks with real-world effects,
• operates with some degree of autonomy.

This standard does not classify intelligence, autonomy level, or behavioral characteristics.

3. Agent Identifier

Each registered agent is assigned a unique identifier in the form:

aiid:00000000000001

The identifier:

• is globally unique,
• is persistent over time,
• does not encode semantic meaning.

Once assigned, an identifier MUST NOT be reused or reassigned.

4. Registry

The registry records:

• agent identifiers,
• basic descriptive metadata,
• declared responsible principals.

The registry does not validate behavior, grant authority, or enforce compliance.

Inclusion in the registry indicates existence only.

Version: Agent ID Standard v0.1
Status: Initial public release