Integrating AI-Based Chatbots for Automated Customer Support in E-Commerce: Using NLP Models like GPT for Real-Time Customer Service Automation

Abstract

This research explores how well AI chatbots operate in online stores using the NLP GPT and especially examining GPT's ability to support immediate customer assistance. Online shopping's quick expansion makes it difficult for stores to give prompt and well-delivered support to their customers. Chatbots make it possible to give user support all day every day at a lower cost with increased user connections. Our systems deliver mixed outcomes when processing complicated and unique customer needs. Our research includes a large quantity of data from 300 customers buying online who used these AI tools to learn about chatbots' trustworthiness and satisfaction ratings alongside correct response evaluation. Users found chatbots to answer simple requests well but struggled with 48% incomplete solutions while being satisfied with chatbot performance at just 31%. The research shows that people base 73% of their trust and recommendation decisions on how accurately and fast a chatbot responds. The research applies TAM to study human acceptance of new technology and shows why people worry about their data privacy in AI systems that struggle to understand emotions. Strong human agent support and visible communication should be integrated alongside continuous NLP training to maintain better chatbot effectiveness. The paper offers useful business methods to improve AI chatbots and prepares the ground for future research on AI systems that handle multiple input types as well as better user feedback testing.