Large Language Models (LLMs) are rapidly evolving and increasingly utilized in a wide range of applications. They offer significant advantages over traditional Natural Language Processing approaches for analyzing requirements written in natural language. When designing intelligent systems according to Industry 4.0 principles, these requirements are often written by stakeholders from different domains, possibly leading to defects and misunderstandings. In this paper, we explore the potential of LLMs for detecting various defects, such as ambiguities, inconsistencies and incompleteness in requirements. We present an approach for improving current requirements engineering processes by defining a conversational interface specifically designed for requirements engineering based on LLMs. This interface allows stakeholders from different domains to interact with requirements without the need for extensive knowledge about the underlying models used by Artificial Intelligence, i.e., mainly the LLMs. This will ultimately provide a bridge between stakeholders and developers to create a common understanding of the requirements. The evaluation results of this approach show that it is possible to detect defects in requirements with a promising level of accuracy and respond with valuable suggestions for improving the quality of requirements.