It is very common in real life that information distorts during the process of transmission in a social network, which may lead to people’s incorrect comprehension of the information and further poor decision making. In this paper, we study how to model and minimize the distortion of information when it diffuses through a social network. We propose the concept of information authenticity to measure distortion as well as a mathematical model to characterize how information distorts during its diffusion through a social network, and study the optimization problem of maximizing the information authenticity of a social network. In order to solve the problem, we employ a framework of greedy algorithms that was proposed by Ni et al. (Inf Sci 180(13):2514–2527, 2010), which can trade off between optimality and complexity. Finally, we perform experiments to show the greedy algorithms can effectively solve the problem we propose.