The Covid-19 post-pandemic educational landscape has intensified the demand for instructional designs that are pedagogically robust, technologically adaptable, and resilient to future disruptions. Despite of this fact, institutional responses to generative Artificial Intelligence (AI) in education frequently emphasize restriction rather than guided integration. This study reports on a Competency-based Education and Training (CBET) course for faculty development grounded in the Technological Pedagogical Content Knowledge (TPACK) framework. Participants redesigned a traditionally delivered lesson into hybrid and digitally resilient formats, culminating in AI-supported lesson plan development. Rather than prohibiting AI use, participants engaged in structured reflective practice documenting AI tools used; prompt strategies, humanisation of outputs, and pedagogical judgement. Data were collected through reflective narratives, survey feedback, artefacts analysis, and defended digital poster presentations evaluated by internal and external experts. Participants demonstrated progressive improvement in lesson alignment, pedagogical coherence, and technological integration across design phases. Reflective data indicated increased metacognitive awareness and professional judgement in AI use. Assessment redesign enabled transparent evaluation of competence, with participants articulating design rationale rather than merely presenting products. Guided integration of AI within a TPACK-aligned CBET framework supports authentic learning, professional judgement, and instructional resilience. Restrictive approaches to AI may undermine these outcomes. Assessment redesign, rather than tool prohibition, emerges as the critical determinant of educational integrity is discussed.