# Agent Chaos

> Economic governance of AI agents for CEOs, CFOs and boards. Agent Chaos is the condition in which AI agents proliferate inside an organization faster than its ability to attribute their value, assign accountable owners, control their cost and retire the ones that do not pay back. Founded by Douglas Siqueira.

Core thesis: AI agents are not tools. They are capital allocation decisions. Economic opacity is the earliest symptom of Agent Chaos.

Agent Chaos is independent: it is not a consultancy, sells no software and has no vendor sponsors. Figures cited on the site come from public research and are linked to their sources.

## Key pages

- [What is Agent Chaos?](https://www.agentchaos.org/en/what-is-agent-chaos): Definition, symptoms, the five capabilities of economic governance, and how the term differs from "chaos testing" and the Northeastern "Agents of Chaos" study.
- [Foundational Statement v1.1](https://www.agentchaos.org/en/foundational-statement): The primary document defining Agent Chaos and economic governance for AI-driven organizations.
- [The Board Questions](https://www.agentchaos.org/en/board-questions): Five questions every CEO and board should ask their team about AI agents — inventory, ownership, cost, outcome and review.
- [AI Agent Cost Calculator](https://www.agentchaos.org/en/agent-cost-calculator): Free calculator for the real monthly cost of an AI agent portfolio — model/API and license spend plus the human review and maintenance time that no invoice shows. Formula: invoice + (agents × review minutes/day ÷ 60 × 21 × reviewer hourly cost) + (agents × maintenance hours/month × engineering hourly cost).
- [Executive Briefs](https://www.agentchaos.org/en/briefs): Short monthly briefs on AI agent ROI, ownership, cost and risk.
- [About](https://www.agentchaos.org/en/about): About the initiative and its founder, Douglas Siqueira.

## Executive Briefs

- [Agent Sprawl: When More Agents Means Less Control](https://www.agentchaos.org/en/briefs/agent-sprawl): AI agent sprawl is not about how many agents you run. It is about running them without a portfolio. The signs, the cost and four decisions that contain it.
- [Who Owns an AI Agent?](https://www.agentchaos.org/en/briefs/who-owns-an-ai-agent): When an AI agent fails, accountability splits across four teams. Why "IT owns it" is the wrong answer, and how to name a single economic owner for every agent.
- [Agents of Chaos: What the Northeastern Study Means for Your Board](https://www.agentchaos.org/en/briefs/agents-of-chaos-study-board): Researchers talked six AI agents into leaking data and wiping an email server. The board-level lesson is not about security tools. It is about ownership.
- [Why Per-Agent ROI Is the Wrong Question](https://www.agentchaos.org/en/briefs/ai-agent-roi-wrong-question): Boards want proof that AI agents pay back. Asking for ROI per agent produces false precision. Measure the portfolio instead — here is how.
- [Shadow AI: The Agents You Pay For and Cannot See](https://www.agentchaos.org/en/briefs/shadow-ai-agents): Employees build AI agents faster than IT can govern them. Shadow AI is not a policy failure — it is an inventory failure. What executives can do about it.

## Languages

- [English](https://www.agentchaos.org/en)
- [Português (Brasil)](https://www.agentchaos.org/pt-br)
- [Español](https://www.agentchaos.org/es)

## Optional

- [RSS feed of Executive Briefs](https://www.agentchaos.org/feed.xml)
- [Full text for LLMs](https://www.agentchaos.org/llms-full.txt): Definition, Foundational Statement summary and FAQ, The Board Questions and every Executive Brief in plain text.
- [Author on LinkedIn](https://www.linkedin.com/in/douglas-siqueira/)