Phishing is an online attack that takes advantage of victims’ technological ignorance, occasionally including a Uniform Resource Locator (URL). Phishing websites are the primary target for attackers seeking to invade individuals’ personal addresses. Many phishing attack detection solutions have been created; however, many have either been ineffective or lack the compact attributes necessary for practical use. To address this issue, innovative techniques are employed, such as Hybrid Deep Learning (DL) Techniques, which combine CNN and LSTM models to accurately comprehend the intrinsic properties of phishing URLs and recognize them. With an overall accuracy of 99.99%, we experimentally proved on standard datasets that the combined identity detection approach greatly enhances phishing detection capacity. It has been demonstrated that the suggested hybrid deep learning technique is useful addition that will provide the phishing prevention ecosystem with better defense measures over rapidly evolving methods used by phishing.