<p>The current development of Artificial Intelligence (AI), despite being dominated in most areas and sectors like healthcare services, education processes, or customer interactions, reflects a range of various potential uses, including conversational AI models such as ChatGPT, Gemini, and Claude. Nevertheless, these models have several potential drawbacks, including the persistence of bias in algorithms in use, various potential ethical questions about user data privacy and security levels, a lack of transparency in the processes of decision-making, and varying accuracy in response to complex questions directed towards these models or a range of other chatbot models in general. This article is based on a systematic review of the technical and scientific literature available between 2019 and 2025 on a range of various potential issues about the overall process of development and evaluation of various models of chatbots in general. This article on the history of chatbot technology surveys the developments in chatbot technology, distinguishing between domain-specific and general-purpose chatbots, and exploring the role played by some of the most important models in establishing contemporary trends. This article also surveys different types of bias and hallucination in the outputs provided by chatbots and discusses their implications for User Interface and User eXperience (UI/UX) design, critically evaluating the varying systems used for evaluating these models. Also, it debates some of the regulatory challenges and compliance issues regarding data usage. Despite its acknowledgement of some of the proposed strategies for mitigation, this review does not attempt to create a mechanized approach for overcoming most, if not all, identified limitations; in fact, it advocates for more rigorous, empirically validated, and user-centric approaches for research of the future.</p>