Microwave heating, which is caused by the interaction of electromagnetic radiation and materials, has become an important component in industrial operations across numerous industries. Despite their importance, conventional numerical simulations of microwave heating are computationally intensive. Concurrently, advances in artificial intelligence (AI), particularly machine learning algorithms, have transformed data processing by increasing accuracy while decreasing computational time. This study tackles the difficulty of efficient and accurate modelling in microwave heating by combining convolutional neural networks (CNNs) with traditional simulation techniques. The major goal of this research is to use CNNs to forecast temperature profiles in a variety of industrial materials, including susceptors, semi-transparent, and microwave-transparent materials, under varying power settings and heating periods. This unique strategy greatly reduces prediction times, with up to 60-fold speed increases over standard methods. Our research is based on examining the electromagnetic and thermal responses of these materials under microwave heating. This study’s findings emphasise the need for extensive datasets and show the transformational potential of CNNs in optimising material processing. It uses artificial intelligence to pave the way for more effective and exact simulations, supporting breakthroughs in industrial microwave heating applications.