Recommender systems predict user preferences and behaviors by analyzing user data, helping businesses enhance customer satisfaction, streamline decision-making, and improve operational efficiency. However, the growing reliance on such systems has raised significant privacy and security concerns. The large-scale collection and analysis of personal data expose users to risks such as data breaches, unauthorized access, and the misuse of sensitive information, which can lead to financial loss, identity theft, and a decline in user trust. While previous research has either focused on applying cryptographic techniques for privacy protection in recommender systems or discussed federated learning in isolation, no comprehensive study has provided a detailed overview of privacy-preserving recommender systems. This review addresses the delicate balance between delivering personalized recommendations and ensuring the privacy and security of user data. It examines existing privacy protection techniques and security measures, highlighting emerging trends and technologies that hold the potential for enhancing privacy and security, including Homomorphic Encryption, Differential Privacy, Federated Learning, and Machine Unlearning. The aim of this review is to provide a thorough understanding of the challenges and solutions involved in creating secure and trustworthy recommender systems, ultimately contributing to the development of more robust and reliable digital services.