The safety of food and beverages has emerged as an urgent concern, as adulterated food and drink can seriously affect human health when consumed. Within this context, the spirits industry stands out as a sector that requires particular attention, as spirits such as whisky, gin, and vodka possess distinct flavor and aroma profiles intrinsically linked to their alcohol composition. To prevent adulteration and ensure the authenticity of spirits, it is essential to identify the purity of specific aroma compounds. In this work, we employed an electronic nose system with an AI algorithm to extract the scent profiles of various whisky spirits, generating their unique aromatic signatures.The AI algorithm demonstrated exceptional performance in classifying different whisky types, achieving an accuracy of 93% to 94%, depending on the type of whisky. Among the models tested, the convolutional neural network-long short-term memory (CNN-LSTM) model consistently outperformed other architectures, including traditional recurrent neural network (RNN), CNN, and LSTM models. The CNN-LSTM model exhibited the highest accuracy and lowest loss, highlighting its superior capability in capturing the complex aroma patterns of whisky spirits. This study represents a significant step forward in ensuring the integrity of spirits; a method for the rapid identification of spirits purity is also provided.