Accurate and prompt diagnosis is crucial for early detection and recovery from breast cancer. Many women die from breast cancer because current methods fail to detect it early. To address this challenge, we designed an Artificial Intelligence-based Random Forest (AI-RF) model optimized with Sail Fish Optimization (SFO). Using MRI breast cancer datasets from Kaggle, the system preprocesses data with median filtering and extracts features using the Grey Level Coherence Matrix (GLCM) method. The SFO fitness is updated in the RF model to improve prediction accuracy of breast cancer stages (benign and malignant). The developed model aims to enhance detection accuracy and reduce false rates. Experimental results show the model achieves 99.13% accuracy, 99.33% sensitivity, and 98.85% precision, outperforming existing techniques.