Lithium-ion batteries prove to be a promising technology for achieving present and future goals regarding energy resources. However, a few cases of lithium-ion battery fires and failures caused by thermal runaway have been reported in various news articles; therefore, it is important to enhance the safety of the batteries and their end users. The early detection of thermal runaway by detecting gases/volatile organic compounds (VOCs) released at the initial stages of thermal runaway can be used as a warning to end users. An interdigitated platinum electrode spin-coated with a sub-micron thick layer of poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT: PSS) showed sensitivity for two VOCs (ethyl-methyl carbonate and methyl formate) released from Li-ion batteries during thermal runaway, as well as their binary mixtures at elevated temperatures, which were measured using impedance spectroscopy over a frequency range of 1 MHz to 1 Hz. The sensor response was tested at three different high temperatures (40 °C, 55 °C, and 70 °C) for single analytes and binary mixtures of two VOCs at 5 ppm, 15 ppm, and 30 ppm concentrations. Equivalent electrical parameters were derived from impedance data. A machine learning approach was used to classify the sensor’s response. Classification algorithms classify the sensor’s response at elevated temperatures for different analytes with an accuracy greater than 70%. The success of the reported sensors will enhance battery safety via the early detection of thermal runaway.