A number of research in Information and Communication Technology (ICT) have shown that usability is an important goal for software-based interactive systems and products for a wider acceptance by their end-users. To evaluate the usability of a system under design, usability evaluation methods predominantly rely on subject matter experts or testers’ assessments. However, integrating the Artificial Intelligence (AI) technology and existing usability evaluation methods’ processes can aid to evolve more effective user-centered designs. In this paper, we conceptualize an AI-augmented usability evaluation framework (AIUEF) that aims at replacing ‘end-users’ with ‘personas’ to evaluate requirements involving human-computer interaction (HCI) for a given Software Requirements Specification (SRS). We present a blueprint of AIUEF that is the first step to furnishing a theoretical basis for automated usability evaluation of HCI requirements for a given SRS with an aim to improve the usability of a system under design.