What is an AI Chatbot vs. Agent?

The distinction is about capability boundary, not conversational tone. A chatbot's job ends once it produces a helpful response; it doesn't independently verify the response accomplished anything in the world, because it typically has no way to act in the world at all. An agent's job doesn't end at a response, it ends when the actual underlying task is complete, which might require several tool calls, checking results, and adjusting course if the first attempt didn't work. This matters practically because the two need very different safeguards. A chatbot's worst failure mode is usually a wrong or unhelpful answer. An agent's worst failure mode can be a wrong real-world action, sending an email to the wrong person, deleting a file, making an unauthorized purchase, which is why agents need sandboxing, tool permissioning, and audit trails that a pure chatbot doesn't require.