The injection molding industry has to face various challenges, such as the shortage of skilled workers and regulatory requirements like the increasing use of recycled materials. These challenges can be overcome, for example, with artificial intelligence (AI) that predicts component quality and suggests machine control parameters. This could drastically reduce the time and effort required for quality assurance, and it would also make it possible to react more quickly to batch fluctuations caused by recycled materials. Various approaches have already been researched. However, the provision of data for the training of such AI models is very extensive and resource-intensive, as the components have to be produced and then measured. Alternatively, simulation data can be used to train the AI models. This publication examines how close the simulation data is to the real data. This enables a better assessment of how good and reliable an AI model could be with simulation data as a basis. Various process and quality data are compared, for example, the cavity pressure and temperature, various dimensions, and the part weight. In certain cases, very good simulative results were achieved that are very close to reality. In addition, the material data stored in the simulation program is also examined to control how reliable the data is for the simulations.