The emergence of online recruitment platforms has greatly changed the hiring landscape, providing easier access to job opportunities. Nevertheless, this convenience has a drawback-an alarming rise in fraudulent job listings that take advantage of job seekers. To tackle this problem, this study introduces a framework driven by AI that autonomously detects and removes fake job listings. The suggested system employs a dataset that includes company-specific details like the company name, license number, review ratings, and legal case history. Natural Language Processing (NLP) methods are utilized to identify semantic patterns within job descriptions, while a Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM) architecture is applied for predictive analysis. The model is designed to determine if job postings are authentic or fake with great precision. Once identified, the system independently removes fraudulent posts, thus safeguarding users and boosting platform trustworthiness. This automated approach enhances the safety of the recruitment process through the use of deep learning and data-informed decision-making.