The rise of generative AI tools such as ChatGPT presents both challenges and opportunities for educators using TBL. While concerns about academic integrity persist, these tools also open new pathways for designing more engaging, reflective, and higher-order learning tasks. This paper presents a set of AI-enhanced Application Exercises that blend the traditional “4S” principles of TBL—Significant problem, Same problem, Specific choice, Simultaneous reporting with creative assignment strategies adapted for an AI-rich environment. 

Drawing on approaches such as prompt competitions, critical revision of AI outputs, visual thinking with mind maps, AI-vs-human comparative tasks and debate-style formats, these exercises engage teams in deeper cognitive and metacognitive work. Each design leverages generative AI not as a shortcut, but as a thinking partner, prompting students to evaluate, critique, and improve machine-generated content while reinforcing discipline-specific reasoning. By reimagining Application Exercises in this way, the paper offers a practical framework for integrating AI into TBL classrooms without compromising pedagogical integrity. Implications for assessment, student engagement, and academic policy are also discussed.