The paper introduces the Discrete Gaussian mechanism for differential privacy. The results can be applied to make sure that the no information is leaked in implementations of differentially private algorithms. The authors are urged to make a better connection to the NeurIPS community.