The emergence of IoT is driving rapid growth in the use of wireless devices around the world. In addition, 5G, through its support for large numbers of devices, is expected to unleash the full potential of IoT globally. The growing number of wireless devices and the massive traffic stemming from the emergence of these technologies will result in a dramatic increase in the demand for spectrum resources. Dynamic spectrum sharing, enabled through cognitive radio network technology, emerges as a key solution for coping with these rising spectrum demands. One important approach that is currently being adopted as a potential solution for promoting dynamic spectrum sharing is the deployment and reliance on white space geo-location spectrum databases for locating spectrum availability in the TV white space. Despite the great benefits these databases offer in terms of their ability to help locate spectrum opportunities for secondary usage, they suffer from location privacy issues, as users need to reveal their location in the process of querying these databases for spectrum availability. Knowing that their whereabouts may be exposed, users can be discouraged from querying the databases, thereby hindering the adoption and deployment of this technology in future generation networks. In this article, we focus on the location privacy problem in database-driven dynamic spectrum access. Specifically, we present and compare key approaches that aim to protect the location information of secondary users in database-driven spectrum sharing and discuss some key research challenges that remain unaddressed.