Cloud adoption in industrial sectors, such as process, manufacturing, health care, and finance, is steadily rising, but as it grows, the risk of targeted cyberattacks has increased. Hence, effectively defending against such attacks necessitates skilled cybersecurity professionals. Traditional human-based cyber-physical education is resource intensive and faces challenges in keeping pace with rapidly evolving technologies. This research focuses on the main advantages of incorporating large language models into cyber-physical education. The ChatGPT platform serves as an online tool to educate students on fundamentals, cyberattacks, and defense concepts, fostering the development of a new generation cybersecurity experts. The proposed learning approach adheres to the ChatGPT-assisted learn-apply-create model. Responding to prompts provided by the learners, the learning phase engages in conceptual learning, the applying phase involves mathematical modeling of various cyberattacks, and the creating phase develops MATLAB program to incorporate attacks into sensor measurements for the experiment and entails developing the necessary attack detection approaches. The effectiveness of the detection method developed by ChatGPT is assessed in both the simulation and real-time scenarios using a J-type thermocouple. The impact of the proposed learning platform over traditional learning methods is evaluated through an extensive comparative feedback analysis on the learner's foundational concepts, computational thinking, programming efficacy, and motivation. The study proved that integrating ChatGPT into engineering education enables students to swiftly learn cyber-physical fundamentals, comprehend and model cyberattacks, create new attack signatures, and contribute to developing detection algorithms. Such integration provides the learners with essential industrial skills crucial in modern industries.