User participation and user feedback are essential to the success of new product development (NPD). Development teams use user feedback to derive requirement engineering artifacts, such as user scenarios, user stories, concept mindmaps, and user personas, to guide them in identifying and addressing a particular user problem. However, finding enough user participation to collect meaningful feedback is challenging, and less attention has been given to addressing this. In this paper, we propose Active Personas (APs), fictional users capable of generating contextual feedback through an interactive multi-modal interaction, such as text, voice, image, and video. APs enable development teams to gather feedback on their solutions through iterative internal experimentation. APs use generative artificial intelligence to enable dynamic multi-modal interaction and utilize user personas to generate contextual feedback. We aim to conduct a series of studies to further validate APs by applying the design science approach as guidance. We plan to develop an initial prototype of APs and conduct studies in a more controlled setting using open-source or completed projects to validate APs. Later, we will transition to ongoing projects in various types and domains.