An AI agent is a system that uses a language model (LLM) to decide and carry out a sequence of actions toward a goal, not just answer a question. It perceives context, picks tools (search the web, call an API, write a file), acts and evaluates the result, repeating the loop until the task is done.
Difference from a plain chatbot
A chatbot answers. An agent acts. The chatbot returns text; the agent can open a ticket, book a slot, generate and run code, or orchestrate other agents. The key word is autonomy with tools: the model stops being a "text generator" and becomes the brain that coordinates real-world actions.
In practice
AI agents already handle support, data analysis, code generation and research. The risk is running them unsupervised: since any AI output can be wrong, the best systems keep a human in the loop and verification at every sensitive step. That control, more than the model itself, separates a reliable agent from an unpredictable one.