In the context of food and economy, meat plays a vital role in fulfilling the nutritional needs of society and serves as a strategic economic commodity. However, the difficulty in distinguishing between beef and pork often leads to fraud by meat traders. Particularly in Indonesia, where the consumption of beef and pork is high, this confusion raises significant concerns, especially since pork is prohibited in the Islamic religion. This research aims to address this issue by applying Artificial Intelligence technology, specifically the Convolutional Neural Network (CNN) deep learning method in classifying images of beef, pork, and mixed meat. The study utilizes a dataset of 410 samples, with 70% used for training and 30% for testing. Testing is conducted using a basic CNN model with hyperparameter analysis such as image size, number of epochs, and batch size. Additionally, the dataset is tested using a comparative architecture, namely the ResNet-50 architecture. The best accuracy rate of the CNN model is 82.20%, achieved with an image size of 75 x 75 pixels, 100 epochs, and a batch size of 64. Testing with the ResNet-50 architecture yields the highest accuracy of 76.14%. Evaluation is performed using a confusion matrix with four categories: Accuracy, Precision, Recall, and F1 Score.