AI Governance Institute
← All news

AI Agent

An AI agent is a system that goes beyond answering questions: it takes a goal, plans a series of steps, and uses tools such as email, databases, web browsers, or payment systems to carry the steps out with limited human involvement. For compliance and risk teams, this matters because each action an agent takes can create real-world consequences, such as sending a message, changing a record, or committing spend, often faster than a person could review it. Governance programs therefore need to define what an agent is permitted to do, who is accountable for its actions, and how its activity is logged and audited.

1 item