When discussing the relevance of Data in our society and the digital future, there is always a concern of security and the uses of our own data. The development of synthetic data to train ML or AI tools seems like a solution to avoid compromising existing real data, but is it really more secure? What is the process of creating such synthetic data?