Governing AI Agents with Democratic ‘Algorithmic Institutions’

Faculty Associate Virgilio Almeida and coauthors Fernando Filgueiras and Ricardo Fabrino Mendonça explore governance structures and forms of institutional oversight to maintain human control over agentic AI. They argue that agentic AI systems pose a fundamental challenge to traditional governance models because they simultaneously function as both institutions and actors. In Tech Policy Press, the authors propose a shift toward a polycentric perspective on AI agent governance that brings humans back into the equation: "Agents that make decisions and execute tasks with operational autonomy make governance a central political problem, perhaps demanding an unprecedented combination of regulations that hold humans accountable for defining the intent of an AI agent and the operation and execution of the task through technology."