<p>Lithium-ion batteries have become a crucial energy source in electric vehicles and energy storage systems. Battery safety and reliability have emerged as significant issues with their increased usage. Failure in non-contact current sensors can lead to inaccuracies in internal current measurement, posing serious risks like overcharging and over-discharging, which may cause fires within battery circuits. To address these challenges, this study proposes an add-on-type current sensor fault diagnosis algorithm based on data analysis using a covariance matrix. The current sensor freezing fault can be captured by determining the internal resistance using the current–voltage relationship. The proposed algorithm enables fault detection without imposing additional load or modifications on the existing battery system. The performance of the proposed current sensor fault diagnosis algorithm was experimentally verified by using two cylindrical battery cells having different life conditions.</p>