Phishing attack is a cybercrime in which an attacker traps the victims by sending fake messages that pretend to have come from a legitimate source. A machine learning approach is one of the prominent methods to identify phishing attacks using conventional methods like support vector machine, random forest, decision tree, neural network, and XG boost. In this research a hybrid detection model that combines Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) to detect phishing URLs, and SVM (Support Vector Machine) for the detection of fraudulent email content. The LSTM-CNN model leverages the sequence processing capability of LSTM and the feature extraction power of CNN, achieving 99.8% accuracy in detecting phishing URLs. The SVM model trained on TF-IDF vectorized email content demonstrates a high accuracy of 98.93% in identifying phishing emails. This combined approach significantly enhances detection performance compared to traditional models by effectively capturing complex patterns in URLs and email texts.