AI-driven dataset creation in mobile forensics usingLLM-based storyboards

The generation of datasets is crucial for training and validation tasks in digital forensics. Currently,data generation and provisioning are primarily performed manually. In the area of mobile forensics,thereareonlyafewtoolsavailablethatfacilitatethedatapopulationandinjectionintophones.Inthisarticle, we present a novel method for automatic data generation using an AI-driven approach. Wethereforedescribeacompletetool-chainfordatasetcreation.Ourfocusisoncreatingadynamicmodel(storyboard) with the help of LLM agents. The generated activity sequences are then automaticallyexecuted on mobile devices. We successfully implemented the proposed approach within the datacreation and injection framework AutoPodMobile(APM) as part of a proof-of-concept study.