Big data is coming with new challenges in security; involve the three aspects of security: (confidentiality, availability, integrity) and privacy. These chal-lenges are due to the characteristics 5V of data in Big data: velocity, variety, volume, value, and veracity. And depend on several level of security: network, data, applica-tion, and authentication. Furthermore, big data is also promising security. The huge amount of data provides a more security information like data logs. Moreover, big da-ta analysis can be applied to security. Many theories for big data security are pro-posed in literature, covering the different aspects of security and privacy. Recently, different schemes and frameworks are introduced to reach high level of security in big data, based on different security theories. In this paper, we discuss different challeng-es in big data security and privacy, and we introduce recent security theories and works in this filed. A comparative study of latest advances in big data security and privacy is presented.