Chatbots can use both AI and Machine Learning, or be powered by simple AI without the added Machine Learning component. There is no one-size-fits-all chatbot and the different types of chatbots operate at different levels of complexity depending on what they are used for. 

Artificial Intelligence (AI) is using programming to simulate human intelligence and creating machines that can make ‘intelligent’ decisions and do tasks that are usually done by humans. Within this broad AI sphere, chatbots are specifically programmed to respond to human inputs in meaningful and useful ways. 

A rules-based chatbot uses simple decision-trees to determine the flow of the interaction with a human. These are usually basic ‘yes/no’, ‘click’ type of responses and are good at straightforward tasks. 

An AI chatbot uses the power of AI to conduct two-way conversations with people using Natural Language Processing technology. These types of chatbots typically use Machine Learning to continually grow and improve in understanding human language and its nuances. They step into the realm of conversational AI, intent recognition, sentiment analysis, deep learning and neural linguistics. 

To learn more, see: 

• How Do AI Chatbots Work: Exploring the Basics  
• AI Chatbots