Governments across the developing world are adopting digital technologies to improve the performance of public sector supply chains. However, empirical evidence regarding the impact of these technologies on performance is mixed. In this paper, we evaluate the impact of using biometric authentication (BA) devices on the performance of India’s food security program. The program aims to deliver subsidized food grains to nearly 160 million low-income households through approximately 500,000 fair price shops (FPSs). Nearly half of those grains do not reach the beneficiaries and are diverted to the open market for private benefit. We estimate the impact of installing BA devices on reducing diversion in more than 3,300 FPSs in the Indian state of Karnataka between 2013 and 2015 using a difference-in-differences approach. We find that BA devices reduce up to 4% of the current diversion at the last mile. We argue that this relatively low impact is due to the presence of a workaround mechanism which results in imperfect monitoring and validate the decrease in diversion mechanism by estimating the differential impact of BA devices. We complement the empirical analysis with a simulation exercise to estimate the additional value that can be unlocked by using the BA devices. We find that the value of better replenishment planning using sales and inventory information captured by BA devices can be significant. The value of better planning can be as high as six times the value of reduction in diversion, depending on the preintervention levels of demand from genuine beneficiaries and diversion at the FPSs.