Agents, workflows, and DAOs
How organizational autonomy relates to agents, automation, multi-agent systems, and blockchain governance.
#Agents and multi-agent systems
An AI agent selects actions using observations, instructions, and available tools. A multi-agent system contains interacting agents. Organizational research adds roles, relationships, responsibilities, and shared purposes to explain how collective behavior is structured. Gaia and MOISE+ are important examples of this organizational perspective.
An agent can be a component of an organization without being an organization itself. Conversely, one agent with durable state could carry several organizational functions. Counting model calls, personas, or processes does not establish organizational autonomy. The stronger question is which continuing functions and obligations the system actually maintains.
#Workflows and organizations
A workflow specifies an arrangement of activities, dependencies, and conditions. It can include AI decisions and be highly adaptive. An organization may rely on many such workflows while also retaining responsibility for priorities, exceptions, resources, and work that was not specified in advance.
For example, a scheduled report generator can execute a valuable workflow. A continuing analytical organization may additionally choose which approved question deserves investigation, allocate effort, check whether a report answers it, and revisit unresolved findings. The distinction depends on demonstrated functions, rather than whether a diagram looks like a pipeline.
#DAO describes a different dimension
In Ethereum’s reference, a decentralized autonomous organization is collectively owned and governed through proposals, voting, and blockchain-based rules. That account emphasizes ownership, governance, and execution of shared rules. It does not require an AI manager or autonomous reasoning.
A DAO can use AI agents; an AI organization can operate without a blockchain. A hybrid system must describe both dimensions: who can legitimately change the rules, and who chooses and performs work under them. Treating DAO, multi-agent system, and autonomous AI organization as synonyms obscures these independent architectural questions.
Sources & context
- Wooldridge, Jennings & Kinny (2000). The Gaia Methodology for Agent-Oriented Analysis and Design
- Hübner, Sichman & Boissier (2002). A Model for the Structural, Functional, and Deontic Specification of Organizations in Multiagent Systems
- Ethereum.org. What is a DAO?
- Buterin (2014). DAOs, DACs, DAs and More: An Incomplete Terminology Guide
Reference edition 0.1 · 2026-10-14. Read the editorial method.