In this paper we investigate if digitized two sided matching platforms have negative social welfare implications when they are used for the solicitation of casual sex, i.e. if they influence the incidence rate of HIV. We further examine to whom these negative externalities accrue. Using a census of 12 million patients who are subjected to a natural experiment in the state of Florida between 2002 and 2006, findings suggest that the largest negative effect accrues to historically at risk populations (i.e. African Americans, men, and the socio-economic lower class) that, ironically, are also on the disadvantaged side of the digital divide. Disturbingly, we find that populations generally considered at lower risk for HIV, and also relatively digitally advantaged, i.e., Caucasians across the entire socio-economic spectrum, are penalized by the presence of the platform. Finally, results show a striking absence of learning and knowledge diffusion about the risks of platform use.