Definition
AI agent
An AI system able to chain multiple actions autonomously to achieve a goal: research, analyze, draft, use tools.
Unlike a chatbot answering a question, an AI agent breaks a goal into steps and executes them: querying a document base, cross-referencing sources, preparing a deliverable, triggering an action in a third-party tool. It loops until the requested result is reached.
The enterprise stake is not technical prowess but control: which actions can the agent take alone, which require human validation? Best practice is to let the agent run autonomously on reversible, internal actions (research, drafts) and to require a human in the loop for any outbound or irreversible action.
Further reading
Sovereign AI, in practice
Multiple AI models run in the EU or on your infrastructure, with your documents connected.