The proliferation of social media platforms has enhanced global communication and facilitated the spread of hate speech, posing a significant challenge to online communities. This study investigates developing and implementing hybrid deep learning models for automatically detecting hate speech across multiple social media platforms. The research aims to improve detection accuracy and efficiency by combining Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks. The study leverages a dataset of approximately 50,000 comments from the “Piers Morgan Uncensored” YouTube channel. Significant improvements in hate speech detection were observed, with the CNN model achieving an accuracy of 99%, representing a substantial improvement over traditional methods. Practical implications include enhanced content moderation capabilities, allowing online platforms to more effectively identify and mitigate hate speech, reducing the burden on human moderators, and thus fostering a safer and more inclusive digital environment.