The swift evolution of the Internet of Things (IoT) has enhanced our daily lives but also increased the IoT’s vulnerability to cyber threats. Cyber adversaries exploit these vulnerabilities using a variety of tactics, necessitating new security measures. As society relies more on cyberspace for activities such as e-commerce, e-healthcare, and social media, it inadvertently provides opportunities for hackers who often employ social engineering techniques. Traditional security tools are insufficient to address the complex and evolving security challenges of today’s IoT landscape. To counter these threats, deep learning (DL) and machine learning (ML) techniques have industrialized as powerful defenses. They can detect patterns, anomalies, and adapt in real time, promising to strengthen IoT security. In this study, we use ML and DL practices, including the Long Short-Term Memory (LSTM) network model, the Deep Neural Network (DNN) model, and the Extra Tree algorithm, to detect phishing websites. We employ a tokenizer to extract key features from URLs, enabling in-depth URL analysis for phishing detection. Our work aims to enhance online security by identifying malicious websites and contribute to safeguarding users against evolving Internet threats.