The need to harness the large quantities of broad and deep data generated across laboratories worldwide, promote global collaboration and data sharing to accelerate research and development and the testing of new therapies and care models for Alzheimer’s and other dementias is today undisputed. The complexity of the dementia challenge and its heterogeneity requires moving beyond the traditional hypothesispredicated scientific approach to the simultaneous assessment of a multitude of factors within big data to discover the unexpected. Capitalising on big data will require, however, a strong effort at several levels. Big data for dementia is not just important for its size, but also for its scope that will go beyond the borders of the health system, requiring data sharing and collaboration among governments, researchers and industry, i.e. linking different communities together. Current research models are however, not well set up for this complexity. While big data networks are proliferating, and the volume and velocity of personal health and other data are rising, barriers still remain with respect to data sharing efforts. Some barriers are of a technical nature, as for issues related to interoperability and standards, storage, the technical infrastructure to allow data sharing. The most significant challenges are, however cultural, legal and ethical and are related to the lack of an open data culture or the disincentives that researchers and scientists face with respect to the disclosure of data. Public policy has a crucial role to play ensuring that framework conditions to promote data sharing are sound and supportive and in setting the conditions for trust and partnerships.